Research protocol - Evaluating data quality in the Netherlands Perinatal Registry (Perined): A data comparison study using hospital records from the IUGR Risk Selection (IRIS) study

Hilde Plomp; Corine Verhoeven; Lilian Peters; Aimée van Dijk; Wes Onland; Ank de Jonge; Jens Henrichs

doi:10.12688/f1000research.150160.2

Home Browse Research protocol - Evaluating data quality in the Netherlands Perinatal...

ALL Metrics

Views

Downloads

Get PDF

Get XML

Export

▬

✚

Study Protocol

Revised

Research protocol - Evaluating data quality in the Netherlands Perinatal Registry (Perined): A data comparison study using hospital records from the IUGR Risk Selection (IRIS) study

[version 2; peer review: 2 approved]

Hilde Plomp^1,2, Corine Verhoeven^1-6, Lilian Peters^1,2,5, [...] Aimée van Dijk⁷, Wes Onland^8,9, Ank de Jonge^1,2,5,9, Jens Henrichs^1,2,5,6

Hilde Plomp^1,2, Corine Verhoeven^1-6, [...] Lilian Peters^1,2,5, Aimée van Dijk⁷, Wes Onland^8,9, Ank de Jonge^1,2,5,9, Jens Henrichs^1,2,5,6

PUBLISHED 09 Jan 2025

Author details Author details

¹ Midwifery Academy Amsterdam Groningen, Inholland University of Applied Sciences, Amsterdam, The Netherlands
² Midwifery Science, Amsterdam University Medical Centres, Amsterdam, The Netherlands
³ Department of Obstetrics and Gynecology, Maxima Medical Centre, Veldhoven, The Netherlands
⁴ Division of Midwifery, School of Health Sciences, University of Nottingham, Nottingham, England, UK
⁵ Department Primary and Long-term Care, University Meidcal Center Groningen, Groningen, The Netherlands
⁶ Amsterdam Public Health, Amsterdam, The Netherlands
⁷ Perined, Utrecht, The Netherlands
⁸ Emma Children's Hospital, Amsterdam University Medical Centres, Amsterdam, The Netherlands
⁹ Reproduction and Development, Amsterdam University Medical Centres, Amsterdam, The Netherlands

Hilde Plomp
Roles: Methodology, Writing – Original Draft Preparation, Writing – Review & Editing

Corine Verhoeven
Roles: Conceptualization, Supervision, Writing – Original Draft Preparation, Writing – Review & Editing

Lilian Peters
Roles: Writing – Review & Editing

Aimée van Dijk
Roles: Writing – Review & Editing

Wes Onland
Roles: Writing – Review & Editing

Ank de Jonge
Roles: Conceptualization, Supervision, Writing – Original Draft Preparation, Writing – Review & Editing

Jens Henrichs
Roles: Conceptualization, Supervision, Writing – Original Draft Preparation, Writing – Review & Editing

OPEN PEER REVIEW

REVIEWER STATUS

Abstract

Background

The quality of registry based studies depends largely on the data accuracy of the registries. The Dutch Perinatal Registry (Perined) is a nationwide database comprising perinatal data digitally provided by different healthcare providers. Perined data are used for comparing outcomes across regions and healthcare institutions as well as for quality analyses and research purposes. However, little is known about the data quality of the Perined database. Therefore, this research protocol depicts our proposed study assessing the quality of Perined data compared to hospital records and case report forms (CRFs) that were part of the IUGR Risk Selection (IRIS) study.

Methods

In the planned comparison study data from Perined and the IRIS Study will be used. The IRIS study was a large cluster-randomized trial investigating the effectiveness of routine third trimester ultrasonography in reducing severe adverse perinatal outcomes among Dutch low-risk pregnant women. A subsample of the IRIS study of neonates being at risk of severe adverse perinatal outcomes and their mothers will be used. Baseline demographic data were collected by midwives from participating women at inclusion (around 22 weeks’ gestation) using CRFs, and in-depth neonatal and maternal clinical data were retrieved from hospital records by trained research assistants. These latter IRIS study data were linked and compared to Perined data. Completeness of Perined data will be calculated for every variable. The reliability will be assessed as the percent of agreement between Perined and hospital record data or the CRF-based data. Additionally, intra-class correlation coefficients will be calculated for continuous variables, and Kappa and ‘Prevalence-and-Bias-Adjusted Kappa’ will be calculated for categorical variables.

Discussion

The results of the planned comparison study will provide users of Perined data insight in its data quality. This will serve as an example of the accuracy of registry based data used in maternal and neonatal care research.

Keywords

Perined, perinatal database, data quality, completeness, reliability, agreement

Corresponding author: Jens Henrichs

Competing interests: No competing interests were disclosed.

Grant information: The author(s) declared that no grants were involved in supporting this work.

Copyright: © 2025 Plomp H et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

How to cite: Plomp H, Verhoeven C, Peters L et al. Research protocol - Evaluating data quality in the Netherlands Perinatal Registry (Perined): A data comparison study using hospital records from the IUGR Risk Selection (IRIS) study [version 2; peer review: 2 approved]. F1000Research 2025, 13:686 (https://doi.org/10.12688/f1000research.150160.2) First published: 24 Jun 2024, 13:686 (https://doi.org/10.12688/f1000research.150160.1) Latest published: 09 Jan 2025, 13:686 (https://doi.org/10.12688/f1000research.150160.2)

Revised Amendments from Version 1

We better clarify that the current paper is a study protocol.
We added some extra information about:
- The linkage of the data.
- The old and new registration approach.
- Anonymizing the data.

See the authors' detailed response to the review by Patricia Lee King
See the authors' detailed response to the review by Elizabeth Kathleen Darling

Introduction

Many countries have a nationwide perinatal database containing data on maternal and neonatal outcomes, such as mode of birth and birth weight.¹^–³ Countries use national perinatal data for monitoring outcomes, quality analyses and research. The data can also be used for comparing results between hospitals, midwifery practices and regions within the country and between countries.²^,⁴^–⁶ For example, EURO-PERISTAT merges aggregated data of all EU members states, Norway, Iceland, Switzerland and the United Kingdom for comparing maternal and neonatal outcomes between the countries.² For these purposes, high quality of the data included in these national databases is essential. Nevertheless, there is a lack of research and insufficient insight into the data quality of these databases.

In 2018, a study was published about the completeness of birth registries worldwide. Phillips et al. concluded that in 2011 only 40% of the total births in the world were registered in a public database, although registration was better in high income countries.¹ In 2019, two Canadian studies were published on the data accuracy of one of two of the Ontario Birth Registries: the BORN (Better Outcomes Registry and Network.⁷^,⁸ Miao et al. compared the BORN birth registry with the general Canadian clinical hospital database (n = 404,439) by evaluating data on key maternal and neonatal outcomes.⁷ Dunn et al. compared data from BORN with data extracted from hospital records (n = 927), and focussed on key outcomes, e.g. labour type and gestational age at birth, but also considered more specific maternal and neonatal outcomes, e.g. ‘pain relief measures during newborn screening’ or ‘serum bilirubin’.⁸ Both studies found an overall good agreement for most key maternal and neonatal outcomes, e.g. almost perfect agreement for gestational age at birth and date of birth.⁷^,⁸ Remarkably, in the study of Miao et al. the lowest Kappa was found for stillbirth or live birth (Kappa 0.74), because of discrepancies in the coding of stillbirths between the data sources.⁷ The study by Dunn et al. using hospital records as comparison found lower agreement for more specific outcomes, e.g. intention to breastfeed and maternal smoking, than for key outcomes.⁸ The BORN birth registry has also been compared to another Canadian birth registry: CIHI-DAD (Canadian Institute for Health Information Discharge Abstract Database).⁹ When lower agreement between variables was found, disagreement could often be explained by differences in coding or differences in definitions of the compared variables.⁹ Another study conducted in Finland in 2002 on the Finnish medical birth registry showed that check-box questions improved data quality, compared to open-ended questions.¹⁰ In addition to research on the accuracy of birth registry databases, studies concerning the data quality of non-perinatal databases showed that the overall completeness, validity, and reliability of national medical databases in high income countries was good and has improved over the years.¹¹^–¹⁵

The planned study focuses on the evaluation of the quality of data obtained from the Netherlands Perinatal Registry (Perined).¹⁶ Such an evaluation is crucial as Perined serves important healthcare-related and scientific purposes. The aim of Perined is to improve the quality of Dutch perinatal care.¹⁶ The database is used for comparing outcomes between regions, midwifery practices and hospitals, for quality analyses to improve perinatal care and research.⁴ The database contains data on pregnancy, birth and neonatal outcomes.⁴ The data are collected by midwives, obstetricians and paediatricians as part of their regular practice, and a selection of items is shared with Perined.¹⁶ In 2020 97% of all Dutch births were registered in Perined.⁴

Several large-scale Dutch studies used Perined data, for example the DELIVER (Data EersteLIjns VERloskunde) study, ABCD (Amsterdam Born Children and their Development) study and the IRIS (IUGR RIsk Selection) study.¹⁷^–¹⁹ These studies all had more than 5000 participants. Data in these studies were collected in different ways and Perined data were used for key maternal and neonatal outcomes. Other studies are completely based on Perined data, e.g. the development of the latest Dutch birthweight charts.²⁰

This illustrates that Perined data are regularly used for research purposes. However, to the best of our knowledge, the data quality of the Perined database has never been assessed in a study. Therefore, the aim of the planned study is to assess the data quality of the Perined database, by assessing the completeness and agreement of the data as compared to perinatal and maternal peripartum variables based on the original hospital records and baseline demographic data from the IUGR Risk Selection (IRIS) study.

Methods

Design

This research protocol depicts a data comparison study concerning a subsample of data based on the IRIS study.¹⁹^,²¹ To assess agreement, data from Perined will be compared with data extracted from hospital records from the participating mothers and/or neonates or with demographic baseline data collected via Case Report Forms (CRFs) at enrolment of participating women (around 22 weeks’ gestation) completed by the midwives together with the participants.¹⁹^,²¹ The data extracted from the hospital records and CRF-based demographics of the IRIS study are used as reference standards. In short, the IRIS study was a multi-centre nationwide stepped wedge cluster randomized trial investigating the effectiveness of routine ultrasound screening in the third trimester of pregnancy in reducing severe adverse perinatal outcomes in low risk pregnancies.¹⁹ Routine third trimester ultrasound screening was compared with standard care. Sixty primary care midwifery practices in the Netherlands and a total of 13,520 women with a low risk singleton pregnancy participated in the IRIS study. The women were enrolled between February 2015 and February 2016. Data of 13,024 (96.3% of 13,520) participating mothers and neonates were linked to Perined data.¹⁹ Data extracted from the hospital records (n=2884) were collected for neonates at risk for severe adverse perinatal outcomes and for mothers with an at risk neonate and/or with indications of peripartum pathology or hospitalization (see section data collection and selection for more details and variable operationalization).

Perined data

Data for this study were collected in 2015 and 2016. In 2016 98% of the Dutch births were registered in Perined.²² Each profession registers a set of variables.²³ Perined links the datasets of each profession and creates one record for all neonates and their mothers ( Figure 1).²⁴ Some variables are uniquely registered by the respective healthcare providers of one profession, whereas others (e.g. type of birth) are registered by healthcare providers of multiple professions. For every variable that is registered by more than one profession, Perined uses a decision tree to determine which value is leading in the Perined record.²⁴ The records used in this study are all records with information from more than one healthcare profession, thus having a complex data structure. In 2015 and 2016 Perined started with a new registration approach for hospitals in the Netherlands. Some hospitals used the new approach, while others still used the old approach. The old and the new approach differ in the variables that are recorded, some variables have been added in the new approach (e.g. length and weight of the pregnant woman), while others have been removed (e.g. method of conception). Also, for some variables answer options have been changed. For the IRIS study, most of the data were provided according to the old approach. The data provided according to the new approach have been recoded by Perined according to the specifications of the old approach.

Figure 1. Perined links datasets and creates one record for all neonates and their mothers.²⁴

Data collection and selection

Perined data and extracted hospital record data for the planned comparison study are selected using a subsample of the IRIS study. Figure 2 shows the used data sources for this study. For the IRIS study hospital records were selected for in-depth data collection based on certain criteria for the Perined or survey data. If the Perined record suggested a case of perinatal death (between 28 weeks pregnancy and 7 days postpartum), a low Apgar score (<4) at five minutes, a birth weight between percentile 2.3 and 5 and neonatal hospitalization for more than three days, or a birth weight < percentile 2.3, data from hospital records were extracted. Hospital records were also selected based on a longitudinal survey; a subsample of women in the IRIS study (n= 1949) took part in this survey.²¹ The survey data were used to identify cases at risk for a severe adverse perinatal outcome. When mothers indicated in the survey that they had a consultation or referral of care to obstetrician-led care, or that their baby visited a paediatrician or neonatologist, they were included for the in-depth data collection from hospital records if they met certain criteria. Maternal indications to be included in the in-depth data collection included maternal hospitalization during pregnancy and/or postpartum or maternal hospitalization for more than 48 hours after birth without medical intervention or an hospitalization of more than 72 hours related to a caesarean section.²⁵ Trained research assistants performed the data extraction. After data extraction from the hospital records, the amount of error in data entry was analysed. Double entry analyses were performed on hospital records of 111 women. The incidence of data entry error was 3.2% overall, 3.7% for maternal data and 2.6% for neonatal data.¹⁹

Figure 2. Data sources of the planned study.

Extracted data from the selected hospital records and CRFs were to those from the Perined records of the same mothers and neonates in order to compare these two types of records. Data managers linked the records by matching several demographic variables (e.g. date of birth of the mother and postal code). Variables will be included in the current study if they are available in both Perined and either the extracted hospital record data or the CRFs. Table 1 and Table 2 show which variables will be compared. The categories of the variables will be created with data driven approaches. We expect that the data for these variables as derived from the respective source can be coded into the categories as shown in Table 1 and Table 2, or are already available in this way. The syntax comprising the respective coding steps will be published, after publication of the final results as planned in the current study. In case a good comparison between certain variables is not possible, they will be removed and not include in the comparative analyses.

Table 1. Example|Descriptive statistics of the variables to be studied based on Perined and the hospital records/CRFs.

Variable¹
Maternal variables (n=)	Perined data source²	Perined n (%)	Missing n	Hospital records or CRFs* n (%)	Missing n
Maternal age (mean (SD))
Maternal ethnicity
Dutch
Other
Parity
0
1
2
3
4
>4
Responsible profession at start of labour
Midwife-led care
Obstetric-led care
Referral of care from midwife-led care to obstetric-led care during labour
Yes
No
Responsible profession at birth
Midwife-led care
Obstetric-led care
Start of labour
Spontaneous
Induction of labour
Planned caesarean section
Type of birth
Spontaneous
Assisted vaginal birth
Planned caesarean section
Unplanned caesarean section
Epidural/spinal analgesia
Yes
No
Augmentation of labour
Yes
No
Postpartum haemorrhage (>1000 cc blood loss)
Yes
No
Episiotomy
Yes
No
Third or fourth degree perinatal trauma
Yes
No

Neonatal variables (n=)	Perined data source¹	Perined n (%)	Missing N	Hospital records n (%)	Missing n
Gestational age at delivery (days) (median (IQR))^b
Sex
Boy
Girl
Inconclusive
Birth weight (grams) (mean (SD))
Apgar score 5 minute postpartum <7
Yes
No
If pH-value is determined, value (mean (SD))
Consultation by the paediatrician and/or hospitalization of the neonate
Yes
No
Perinatal death
No
Antenatal death
Natal death
Neonatal death (up to 7 days postpartum)

1 Final choices for the categorization per variable will be made when doing the analysis. An attempt will be made to stay as close as possible to the Perined categories as used in the Perined data.

2 As discussed in the methods section Perined variables can be registered by one healthcare profession, but also by more than one. This column will show the data source(s) of the variable.

Table 2. Example|Degree of completeness and the agreement between Perined and hospital records/CRFs to be studied.

Variable
Maternal and birth variables	Degree of completeness Perined (%)	Percent agreement (%)	Kappa (95% CI) or ICC* (95% CI)	PABAK	Cases (n)
Maternal age
Maternal ethnicity
Parity
Responsible profession at start of labour
Referral of care from midwife-led care to obstetric-led care during labour
Responsible profession at birth
Start of labour
Type of birth
Epidural/spinal analgesia
Augmentation of labour
Postpartum haemorrhage (>1000 cc blood loss)
Episiotomy
Third or fourth degree perinatal trauma
*Neonatal variables*
Gestational age at birth (in days)
Sex
Birth weight (in grams)
Apgar score 5 minutes postpartum <7
pH-value umbilical cord determined
If pH-value is determined. value
Consultation by the paediatrician and/or hospitalization of the neonate
Perinatal death
No
Antenatal death
Natal death
Neonatal death (up to 7 days postpartum)

Records with missing data for some of the variables will be included in the planned study, to assess the degree of completeness of the different variables in Perined.

Statistical analyses

The data in the IRIS study are not always coded in the same way as those in Perined. Recoding of the variables will be necessary to make comparisons possible, this will be done in Rstudio 4.2.1 and IBM SPSS statistics 28. We will stay as close as possible to the Perined categories. If too much recoding is required to compare the variables, no comparison will be considered. All statistical analyses will be performed in Rstudio 4.2.1. PABAK will be calculated in Rstudio 4.2.1 with the formula 2*((a+d)/(a+b+c+d))-1, based on a 2x2 table.²⁶ For ordinal scales, PABAK will be calculated using the PABAK-OS calculator.²⁷

By using descriptive statistics, frequencies and percentages will be presented for all categorical variables. For continuous variables (e.g. birthweight, Apgar score) with a normal distribution, means and standard deviations (SD) will be calculated. Medians and interquartile ranges (IQR) will be reported for continuous variables not normally distributed.

The completeness of the selected variables in Perined will be calculated as the number of patients with information recorded in Perined divided by the total number of patients in the hospital records dataset or CRF based dataset.

The agreement of the categorical variables will be assessed as the percent agreement and Cohen’s Kappa with 95% confidence interval (CI). Cohen’s Kappa is a statistical measure for dichotomous, categorical and nominal variables to examine the proportion of agreement corrected for the proportion of agreement that could be expected by chance.⁸^,²⁸ The following criteria to assess the strength of agreement will be used: Kappa coefficient ≤ 0 poor, 0.01-0.20 slight, 0.21-0.40 fair, 0.41-0.60 moderate, 0.61-0.8 substantial, 0.81-0.99 almost perfect.²⁹^,³⁰ The advantage of using Kappa as a measure of agreement is the correction for chance. The disadvantage of using Kappa is the extreme sensitivity for unequal distributions. When one of the values is very common and the other very rare, Kappa is drastically lowered, even when the agreement is high, this is defined as the Kappa’s paradox.³¹ Therefore, the ‘Prevalence-Adjusted-Bias-Adjusted-Kappa’ (PABAK) will also be calculated for all categorical variables as this measure can account for unequal distributions of the test variables.²⁶ For PABAK, the same strength of agreement criteria as for KAPPA will be used.

For continuous variables, the agreement between Perined and the extracted hospital record data will be assessed using the percent agreement and intra-class correlation coefficient (ICC) with 95% CI. The ICC is a correlation coefficient which tests whether there is agreement between two continuous variables measuring the same outcomes in two different data sources. The ICC is an appropriate statistical measure of reliability for continuous data.³² ICC values range between 0 (no agreement) and 1 (total agreement). The following criteria to assess the level of agreement will be used: < 0.5 poor, 0.5-0.75 moderate, 0.75-0.9 good, > 0.90 excellent.³²

Conclusion and discussion

The planned study with data of the IRIS study provide a unique opportunity to validate the Perined dataset on a large scale. In recent years the Dutch Perined database has been increasingly used for various purposes, e.g. benchmarking between Dutch regions, benchmarking between nations, and various types of research. However, so far, little is known about the data quality of Perined despite its importance for healthcare evaluation and research purposes.

The planned study will have some strengths and limitations. The greatest strength of this study is that using a large-scale nationally representative low risk pregnancy population in the Netherlands provides the rare opportunity to validate the quality of Perined data. A potential limitation is that the data collection for the IRIS study was conducted in the years 2015 and 2016. These years were so-called ‘transition years’ for the Dutch perinatal registration during which Perined implemented a new dataset system. For the IRIS study the ‘old’ way of data registration and coding were used. Data from participating care institutions that already used the new approach were recoded to fit in the old dataset format as depicted above. A priori one might assume that this may have influenced the degree of missingness in some perinatal variables. The linkage rate between Perined and the complete IRIS study data was very high (96.3%) as compared to a recent study with a much lower linkage rate (78.5%) using the ‘new data registration approach’ only.¹⁹^,³³ Data used for this study were mostly based on complex records reported by different healthcare professionals, see Figure 1. In several instances, information on a certain variable, e.g. birth weight, was entered by several healthcare professionals so that a decision rule was used by Perined determining from which healthcare profession data should be used for a specific variable. One might assume that the accuracy of variables registered by only one healthcare profession may be more accurate than for those variables entered by multiple professions. However, this is unavoidable as for a large group of women and their offspring in ante- and postnatal care referred from midwifery-led care to obstetrician-led care data is generally entered into the Perined database by multiple healthcare professions.⁴

By investigating the data quality of Perined using data from the IRIS study the planned comparison study can inform researchers about the reliability and usefulness of the Perined-based neonatal and maternal variables addressed in this study. This information will help researchers to make adequate choices for Perined derived outcome variables to be used in future studies. Next to this the results of the current study will help medical software developers to improve data-extraction procedures, and Perined to improve data-collection procedures. In addition, the results of this study can illustrate how national birth registries can be improved to enhance quality improvement activities in perinatal care research.

Ethical considerations

This study is a secondary analysis of the IRIS study which was approved by the Dutch Institutional Review Board of the VU Medical University Centre Amsterdam on the 17^th of December 2013 (reference number 2013.409). This study was performed in line with the ethical principles established by the World Medical Association in the Declaration of Helsinki of Ethical Principles for Medical Research Involving Human Subjects.³⁴ Women gave permission to link their records, the linking was performed by data managers of the IRIS study with the support of Perined. After linking the data all identifiable patient information about patients was removed. Analyses for the planned study will be done with a completely anonymous data set, without any identifying variables (e.g. patient names, date of births and study ID).

Consent to participate

Between February 2015 and February 2016, pregnant women in participating midwifery practices who fulfilled the inclusion criteria were informed about the study and received an information leaflet from their midwife during the first consultation. All participants gave written informed consent for data usage.

Author contributions

Plomp H: Methodology, Writing – Original Draft Preparation, Writing – Review & Editing; Verhoeven CJM, de Jonge A, Henrichs J: Conceptualization, Supervision, Writing – Original Draft Preparation, Writing – Review & Editing. Peters L, van Dijk AE, Onland W: Writing – Review & Editing

Supplementary material

Not applicable.

Data availability

The existing ethical approval for the IRIS study does not permit publication of individual participant level data. Requests for a de-identified dataset require a Data Transfer Agreement that is in line with the European Union’s General Data Protection Regulation (GDPR) and can be sent to the corresponding author (j.henrichs@amsterdamumc.nl).

Software availability

All software code that will be used for this study based on IBM SPSS statistics 28 and Rstudio 4.2.1 can also be made available by the coordinator of the IRIS study.

Rstudio 4.2.1 is open access.³⁵ IBM SPSS statistics 28 is proprietary software, PSPP is a non-proprietary alternative.³⁶^,³⁷

National perinatal registration data from the Netherlands is available publicly from Perined at www.peristat.nl.

Acknowledgements

We thank the participating women, midwives, obstetricians, sonographers, IRIS study data management and research assistants team, and staff at participating midwifery practices and hospitals in the Netherlands and the Perinatal Registry of the Netherlands (Perined) for the use of their database.

References

1. Phillips DE, Adair T, Lopez AD: How useful are registered birth statistics for health and social policy? A global systematic assessment of the availability and quality of birth registration data. Popul. Health Metrics. 2018; 16(1): 21. PubMed Abstract | Publisher Full Text | Free Full Text
2. Project Euro-Peristat: European Perinatal Health Report – Core indicators of the and care of pregnant women and babies in Europe from 2015 to 2019. Paris: Euro-Peristat; 2022.
3. Delnord M, Szamotulska K, Hindori-Mohangoo AD, et al.: Linking databases on perinatal health: a review of the literature and current practices in Europe. Eur. J. Pub. Health. 2016; 26(3): 422–430. PubMed Abstract | Publisher Full Text | Free Full Text
4. Perined: Perinatale zorg in Nederland anno 2020 [Perinatal care in the Netherlands in 2020]. Utrecht: Perined; 2021.
5. Devlieger RME, Goemaes R, Cammu H: Perinatale activiteiten in Vlaanderen 2017 [Perinatal activities in Flanders 2017]. Brussel: SPE; 2018.
6. Draper ESGI, Smith LK, Kurinczuk JJ, et al.: MBRRACE-UK Perinatal Mortality Surveillance Report UK Perinatal Deaths for Births from January to December 2017. Leicester: MBRRACE-UK collaboration; 2019.
7. Miao Q, Fell DB, Dunn S, et al.: Agreement assessment of key maternal and newborn data elements between birth registry and Clinical Administrative Hospital Databases in Ontario, Canada. Arch. Gynecol. Obstet. 2019; 300(1): 135–143. PubMed Abstract | Publisher Full Text
8. Dunn SLA, Sprague AE, Fell DB, et al.: Data accuracy in the Ontario birth Registry: a chart re-abstraction study. BMC Health Serv. Res. 2019; 19: 1001–1011. PubMed Abstract | Publisher Full Text | Free Full Text
9. Darling EK, Marquez O, Park AL, et al.: Defining a low-risk birth cohort: a cohort study comparing two perinatal data sets in Ontario, Canada. Int. J. Popul. Data Sci. 2024; 9(1): 2364. PubMed Abstract | Publisher Full Text | Free Full Text
10. Gissler M, Shelley J: Quality of data on subsequent events in a routine Medical Birth Register. Med. Inform. Internet Med. 2002; 27(1): 33–38. PubMed Abstract | Publisher Full Text
11. Ballantine KR, Hanna S, Macfarlane S, et al.: Childhood cancer registration in New Zealand: A registry collaboration to assess and improve data quality. Cancer Epidemiol. 2018; 55: 104–109. PubMed Abstract | Publisher Full Text
12. Linder G, Lindblad M, Djerf P, et al.: Validation of data quality in the Swedish National Register for Oesophageal and Gastric Cancer. Br. J. Surg. 2016; 103(10): 1326–1335. PubMed Abstract | Publisher Full Text
13. Londero SC, Mathiesen JS, Krogdahl A, et al.: Completeness and validity in a national clinical thyroid cancer database: DATHYRCA. Cancer Epidemiol. 2014; 38(5): 633–637. PubMed Abstract | Publisher Full Text
14. Ostgard LS, Norgaard JM, Severinsen MT, et al.: Data quality in the Danish National Acute Leukemia Registry: a hematological data resource. Clin. Epidemiol. 2013; 5: 335–344. PubMed Abstract | Publisher Full Text | Free Full Text
15. Lofgren L, Eloranta S, Krawiec K, et al.: Validation of data quality in the Swedish National Register for Breast Cancer. BMC Public Health. 2019; 19(1): 495. PubMed Abstract | Publisher Full Text | Free Full Text
16. Perined: Over Perined [About Perined]. Website Perined. Accessed January 23, 2023. Reference Source
17. Mannien J, Klomp T, Wiegers T, et al.: Evaluation of primary care midwifery in The Netherlands: design and rationale of a dynamic cohort study (DELIVER). BMC Health Serv. Res. 2012; 12: 69. PubMed Abstract | Publisher Full Text | Free Full Text
18. Van Eijsden M, Vrijkotte TG, Gemke RJ, et al.: Cohort profile: the Amsterdam Born Children and their Development (ABCD) study. Int. J. Epidemiol. 2011; 40(5): 1176–1186. PubMed Abstract | Publisher Full Text
19. Henrichs J, Verfaille V, Jellema P, et al.: Effectiveness of routine third trimester ultrasonography to reduce adverse perinatal outcomes in low risk pregnancy (the IRIS study): nationwide, pragmatic, multicentre, stepped wedge cluster randomised trial. BMJ. 2019; 367: l5517. PubMed Abstract | Publisher Full Text | Free Full Text
20. Hoftiezer L, Hof MPH, Dijs-Elsinga J, et al.: From population reference to national standard: new and improved birthweight charts. Am. J. Obstet. Gynecol. 2019; 220(4): 383.e1–383.e17. PubMed Abstract | Publisher Full Text
21. Henrichs J, Verfaille V, Viester L, et al.: Effectiveness and cost-effectiveness of routine third trimester ultrasound screening for intrauterine growth restriction: study protocol of a nationwide stepped wedge cluster-randomized trial in The Netherlands (The IRIS Study). BMC Pregnancy Childbirth. 2016; 16(1): 310. Publisher Full Text
22. Perined: Perinatale Zorg in Nederland 2016 [Perinatale Care in the Netherlands 2016]. Utrecht: Perined; 2018.
23. Perined: Registratie [Registration]. Website Perined. Accessed January 23, 2023. Reference Source
24. Perined: Methodologie Perinatologie Data: Van ontvangst tot gebruiksklare data voor onderzoek, benchmarking en rapportage [Methodology Perinatology Data: From receipt to ready-to-use data for research, benchmarking and reporting]. Utrecht: 2020.
25. Henrichs J, De Jonge A, Westerneng M, et al.: Cost-Effectiveness of Routine Third Trimester Ultrasound Screening for Fetal Growth Restriction Compared to Care as Usual in Low-Risk Pregnancies: A Pragmatic Nationwide Stepped-Wedge Cluster-Randomized Trial in The Netherlands (the IRIS Study). Int. J. Environ. Res. Public Health. 2022; 19(3312). PubMed Abstract | Publisher Full Text | Free Full Text
26. Byrt T, Bishop J, Carlin JB: Bias, prevalence and kappa. J. Clin. Epidemiol. 1993; 46(5): 423–429. Publisher Full Text
27. Vannest KJ, Parker RI, Gonen O, et al.: Single Case Research: web based calculators for SCR analysis. (Version 2.0) [Web-based application]. Accessed February 7, 2024. Reference Source
28. Twisk J: Inleiding in de toegepaste biostatistiek [Introduction in applied biostatistics]. Houten: Bohn Stafleu van Loghum; 2017.
29. Landis JR, Koch GG: The measurement of observer agreement for categorical data. Biometrics. 1977; 33(1): 159–174. Publisher Full Text
30. Viera AJ, Garrett JM: Understanding interobserver agreement: the kappa statistic. Fam. Med. 2005; 37(5): 360–363. PubMed Abstract
31. Feinstein AR, Cicchetti DV: High agreement but low kappa: I. The problems of two paradoxes. J. Clin. Epidemiol. 1990; 43(6): 543–549. PubMed Abstract | Publisher Full Text
32. Koo TK, Li MY: A Guideline of Selecting and Reporting Intraclass Correlation Coefficients for Reliability Research. J. Chiropr. Med. 2016; 15(2): 155–163. PubMed Abstract | Publisher Full Text | Free Full Text
33. Becking EC, Scheffer PG, Henrichs J, et al.: Fetal fraction of cell-free DNA in noninvasive prenatal testing and adverse pregnancy outcomes: a nationwide retrospective cohort study of 56,110 pregnant women. Am. J. Obstet. Gynecol. 2023. PubMed Abstract | Publisher Full Text
34. The Helsinki Declaration of the World Medical Association (WMA): Ethical principles of medical research involving human subjects. Pol Merkur Lekarski. 2014; 36: 301. Publisher Full Text
35. RStudio Team: RStudio: Integrated Development for R. Boston, MA: RStudio; 2020.
36. IBM Corp: IBM SPSS Statistics for Windows, Version 28.0. Armonk, NY: IBM Corp; 2021.
37. Pfaff B, Darrington J, Stover J, et al.: GNU PSPP Statistical Analysis Software: Release 0.9. Boston, MA: Free Software Foundation; 2007.

Comments on this article Comments (0)

Version 2

VERSION 2 PUBLISHED 24 Jun 2024

Author details Author details

Hilde Plomp
Roles: Methodology, Writing – Original Draft Preparation, Writing – Review & Editing

Corine Verhoeven
Roles: Conceptualization, Supervision, Writing – Original Draft Preparation, Writing – Review & Editing

Lilian Peters
Roles: Writing – Review & Editing

Aimée van Dijk
Roles: Writing – Review & Editing

Wes Onland
Roles: Writing – Review & Editing

Ank de Jonge
Roles: Conceptualization, Supervision, Writing – Original Draft Preparation, Writing – Review & Editing

Jens Henrichs
Roles: Conceptualization, Supervision, Writing – Original Draft Preparation, Writing – Review & Editing

Competing interests

No competing interests were disclosed.

Grant information

The author(s) declared that no grants were involved in supporting this work.

Article Versions (2)

version 2

Revised

Published: 09 Jan 2025, 13:686

https://doi.org/10.12688/f1000research.150160.2

version 1

Published: 24 Jun 2024, 13:686

https://doi.org/10.12688/f1000research.150160.1

© 2025 Plomp H et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Download

Export To

metrics

	Views	Downloads
F1000Research	-	-
PubMed Central Data from PMC are received and updated monthly.	-	-

Citations

SEE MORE DETAILS

CITE

how to cite this article

Plomp H, Verhoeven C, Peters L et al. Research protocol - Evaluating data quality in the Netherlands Perinatal Registry (Perined): A data comparison study using hospital records from the IUGR Risk Selection (IRIS) study [version 2; peer review: 2 approved]. F1000Research 2025, 13:686 (https://doi.org/10.12688/f1000research.150160.2)

NOTE: If applicable, it is important to ensure the information in square brackets after the title is included in all citations of this article.

track

receive updates on this article

Track an article to receive email alerts on any updates to this article.

Open Peer Review

Version 2

VERSION 2

PUBLISHED 09 Jan 2025

Revised

Views

Reviewer Report 12 Feb 2025

Elizabeth Kathleen Darling, McMaster University, Hamilton, Ontario, Canada

Approved

https://doi.org/10.5256/f1000research.175292.r357737

Thank you for your thoughtful and detailed responses to my previous ... Continue reading

CITE

Report a concern

Respond or Comment

Views

Reviewer Report 03 Feb 2025

Patricia Lee King, University of Chicago Pritzker School of Medicine, Chicago, USA; Medical Social Sciences, Northwestern University - Chicago (Ringgold ID: 205058), Chicago, Illinois, USA

Approved

https://doi.org/10.5256/f1000research.175292.r357738

CITE

Report a concern

Respond or Comment

Version 1

VERSION 1

PUBLISHED 24 Jun 2024

Views

Reviewer Report 16 Jul 2024

Elizabeth Kathleen Darling, McMaster University, Hamilton, Ontario, Canada

Approved with Reservations

https://doi.org/10.5256/f1000research.164702.r295929

This manuscript presents a study protocol for a data quality study that will compare data collected in the Netherlands Perinatal Registry (Perined) with data collected via hospital records and case report forms for participants in a large cluster-randomized trial (the IRIS study). The authors make a strong case for the importance of high-quality national perinatal registry data to support quality improvement and research, and they successfully argue that the use of high-quality data collected as part of an RCT presents a unique opportunity to assess the quality of Perined data.

The objectives of the study are clearly stated and the study design is appropriate to address the objectives. The data sources and the methods originally used to collect the data are adequately described. The plans for statistical analysis are appropriate and are well described. The dummy tables that are provided (Tables 1 & 2) are appropriate to present the planned analyses. (Revising the titles of these tables might help to make their purpose more clear to the reader.)

The protocol could be strengthened by the addition of some key information. First, no details are provided regarding how the data sources will be linked. The authors cite a prior study that linked Perined and IRIS data, implicitly demonstrating feasibility, but do not explain what variable(s) will be used for linkage. Furthermore, under 'Ethical considerations', the authors state that the analyses will be done using an anonymous data source, without patient names and study ID, which again raises the question of how the data will be linked. If a previously linked data set is being used, this should be explained. Second, the protocol for the study should include the algorithms that will be used to recode some variables to allow for comparison between two sources. This could be presented as the first table (or a supplementary table if it is very lengthy), and should ideally include the name of the variable in each source and the rules that will be applied to to recode the comparator variables to align them with Perined variables when applicable. This level of detail is necessary to permit replication of the study. Finally, it would be helpful for the reader if you could add another sentence or two to explain the difference between the old and new registration approaches for hospitals that are/were used for Perined. This additional contextual detail would help the reader understand the potential implications of the change in approach.

Given the citation of two similar studies in the introduction, the authors may be interested in this recent publication that also examines data accuracy of the Ontario Birth Registry (BORN) -Ref [1]

I look forward to reading the findings from this study once it is completed.

Is the rationale for, and objectives of, the study clearly described?

Yes
Is the study design appropriate for the research question?

Yes
Are sufficient details of the methods provided to allow replication by others?

No
Are the datasets clearly presented in a useable and accessible format?

Not applicable

References

1. Darling EK, Marquez O, Park AL: Defining a low-risk birth cohort: a cohort study comparing two perinatal data sets in Ontario, Canada.Int J Popul Data Sci. 2024; 9 (1): 2364 PubMed Abstract | Publisher Full Text

Competing Interests: No competing interests were disclosed.

Reviewer Expertise: midwifery services, large database research, access to care, health equity

I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.

CITE

Report a concern

Author Response 09 Jan 2025

Hilde Plomp, Midwifery Science, Amsterdam University Medical Centres, Amsterdam, The Netherlands

09 Jan 2025

Author Response

Dear Elizabeth K. Darling,

We would like to thank you for reading and reviewing the article, for the positive evaluation and compliments concerning our article, and for the additional ... Continue reading Dear Elizabeth K. Darling,

We would like to thank you for reading and reviewing the article, for the positive evaluation and compliments concerning our article, and for the additional thoughtful comments and suggestions. Below is a point-by-point reply to your comments. We indicated the start of your comments by using your initials (EKD) and our respective reply starts with HP (i.e. Hilde Plomp the first author answering on behalf of all the authors). Changed text passages, which are presented below, are indicated by using quotation marks, and for the convenience of the reviewer, we also refer to the respective pages and lines in the manuscript where these passages can be found.

Reviewer: Revising the titles of these tables might help to make their purpose more clear to the reader.
Author Response: We have changed the titles of the tables to make it more clear to the reader that these are example tables and what they contain. For example, the title of Table 1 now reads:
“Table 1. Example|Descriptive statistics of the variables to be studied based on Perined and the hospital records/CRFs.”

Reviewer: First, no details are provided regarding how the data sources will be linked. The authors cite a prior study that linked Perined and IRIS data, implicitly demonstrating feasibility, but do not explain what variable(s) will be used for linkage.”
Author Response: We have added some additional information about the data linkage procedure in the Methods section, as shown below (see also page 6 of the manuscript):
“Extracted data from the selected hospital records and CRFs were linked to those from the Perined records of the same mothers and neonates in order to compare these two types of records. Data managers linked the records by matching several demographic variables (e.g. date of birth of the mother and postal code). Variables chosen for analysis will be included in the planned study if they are available in both Perined and either the extracted hospital record data or the CRFs.”

Reviewer: Furthermore, under 'Ethical considerations', the authors state that the analyses will be done using an anonymous data source, without patient names and study ID, which again raises the question of how the data will be linked. If a previously linked data set is being used, this should be explained.
Author Response: We have added some extra information in the ‘Ethical considerations’ section. It is important to note that the datasets were not completely anonymous to the data managers involved in our study. However, researchers receive anonymized datasets for data-analysis purposes. Our additions to the ‘Ethical considerations’ section are shown below (see also page 14):
“Women gave permission to link their records, the linking was performed by data managers of the IRIS study with the support of Perined. After linking the data all identifiable patient information about patients was removed. Analyses for the planned study will be done with a completely anonymous data set, without any identifying variables (e.g. patient names, date of births and study ID).”

Reviewer: Second, the protocol for the study should include the algorithms that will be used to recode some variables to allow for comparison between two sources. This could be presented as the first table (or a supplementary table if it is very lengthy), and should ideally include the name of the variable in each source and the rules that will be applied to recode the comparator variables to align them with Perined variables when applicable. This level of detail is necessary to permit replication of the study.
Author Response: We cannot include the requested syntax in the current article as the stage of our planned work does not allow this. The syntax of the recoding will be published as a supplementary file included in the subsequent scientific article reporting our final results. Until we have completed the variable coding and data-analysis, we cannot guarantee full precision regarding the best recoding method for each variable (see also page 6):
“The categories of the variables will be created with data driven approaches. We expect that the data for these variables as derived from the respective source can be coded into the categories as shown in Tables 1 and 2, or are already available in this way. The syntax comprising the respective coding steps will be published, after publication of the final results as planned in the current study. In case a good comparison between certain variables is not possible, they will be removed and not included in the comparative analyses.”

Reviewer: Finally, it would be helpful for the reader if you could add another sentence or two to explain the difference between the old and new registration approaches for hospitals that are/were used for Perined. This additional contextual detail would help the reader understand the potential implications of the change in approach.
Author Response: We have added some additional information on this aspect in the Methods section (see also page 5):
“In 2015 and 2016 Perined started with a new registration approach for hospitals in the Netherlands. Some hospitals used the new approach, while others still used the old approach. The old and the new approach differ in the variables that are recorded, some variables have been added in the new approach, while others have been removed. Also, for some variables answer options have been changed. For the IRIS study, most of the data were provided according to the old approach. The data provided according to the new approach have been recoded by Perined according to the specifications of the old approach.”

Reviewer: Given the citation of two similar studies in the introduction, the authors may be interested in this recent publication that also examines data accuracy of the Ontario Birth Registry (BORN) -Ref [1].
Author Response: thank you for the suggestion. We have added your recent publication in the Introduction (see also page 4).
“The BORN birth registry has also been compared to another Canadian birth registry: CIHI-DAD (Canadian Institute for Health Information Discharge Abstract Database) [Darlings et al., 2024]. When lower agreement between variables was found, disagreement could often be explained by differences in coding or differences in definitions of the compared variables [bron: Darlings, 2024].”

We hope that our changes are satisfactory. We are looking forward to your evaluation.

Kind regards, on behalf of all authors,

Hilde Plomp
Dear Elizabeth K. Darling,

We would like to thank you for reading and reviewing the article, for the positive evaluation and compliments concerning our article, and for the additional thoughtful comments and suggestions. Below is a point-by-point reply to your comments. We indicated the start of your comments by using your initials (EKD) and our respective reply starts with HP (i.e. Hilde Plomp the first author answering on behalf of all the authors). Changed text passages, which are presented below, are indicated by using quotation marks, and for the convenience of the reviewer, we also refer to the respective pages and lines in the manuscript where these passages can be found.

Reviewer: Revising the titles of these tables might help to make their purpose more clear to the reader.
Author Response: We have changed the titles of the tables to make it more clear to the reader that these are example tables and what they contain. For example, the title of Table 1 now reads:
“Table 1. Example|Descriptive statistics of the variables to be studied based on Perined and the hospital records/CRFs.”

Reviewer: First, no details are provided regarding how the data sources will be linked. The authors cite a prior study that linked Perined and IRIS data, implicitly demonstrating feasibility, but do not explain what variable(s) will be used for linkage.”
Author Response: We have added some additional information about the data linkage procedure in the Methods section, as shown below (see also page 6 of the manuscript):
“Extracted data from the selected hospital records and CRFs were linked to those from the Perined records of the same mothers and neonates in order to compare these two types of records. Data managers linked the records by matching several demographic variables (e.g. date of birth of the mother and postal code). Variables chosen for analysis will be included in the planned study if they are available in both Perined and either the extracted hospital record data or the CRFs.”

Reviewer: Furthermore, under 'Ethical considerations', the authors state that the analyses will be done using an anonymous data source, without patient names and study ID, which again raises the question of how the data will be linked. If a previously linked data set is being used, this should be explained.
Author Response: We have added some extra information in the ‘Ethical considerations’ section. It is important to note that the datasets were not completely anonymous to the data managers involved in our study. However, researchers receive anonymized datasets for data-analysis purposes. Our additions to the ‘Ethical considerations’ section are shown below (see also page 14):
“Women gave permission to link their records, the linking was performed by data managers of the IRIS study with the support of Perined. After linking the data all identifiable patient information about patients was removed. Analyses for the planned study will be done with a completely anonymous data set, without any identifying variables (e.g. patient names, date of births and study ID).”

Reviewer: Second, the protocol for the study should include the algorithms that will be used to recode some variables to allow for comparison between two sources. This could be presented as the first table (or a supplementary table if it is very lengthy), and should ideally include the name of the variable in each source and the rules that will be applied to recode the comparator variables to align them with Perined variables when applicable. This level of detail is necessary to permit replication of the study.
Author Response: We cannot include the requested syntax in the current article as the stage of our planned work does not allow this. The syntax of the recoding will be published as a supplementary file included in the subsequent scientific article reporting our final results. Until we have completed the variable coding and data-analysis, we cannot guarantee full precision regarding the best recoding method for each variable (see also page 6):
“The categories of the variables will be created with data driven approaches. We expect that the data for these variables as derived from the respective source can be coded into the categories as shown in Tables 1 and 2, or are already available in this way. The syntax comprising the respective coding steps will be published, after publication of the final results as planned in the current study. In case a good comparison between certain variables is not possible, they will be removed and not included in the comparative analyses.”

Reviewer: Finally, it would be helpful for the reader if you could add another sentence or two to explain the difference between the old and new registration approaches for hospitals that are/were used for Perined. This additional contextual detail would help the reader understand the potential implications of the change in approach.
Author Response: We have added some additional information on this aspect in the Methods section (see also page 5):
“In 2015 and 2016 Perined started with a new registration approach for hospitals in the Netherlands. Some hospitals used the new approach, while others still used the old approach. The old and the new approach differ in the variables that are recorded, some variables have been added in the new approach, while others have been removed. Also, for some variables answer options have been changed. For the IRIS study, most of the data were provided according to the old approach. The data provided according to the new approach have been recoded by Perined according to the specifications of the old approach.”

Reviewer: Given the citation of two similar studies in the introduction, the authors may be interested in this recent publication that also examines data accuracy of the Ontario Birth Registry (BORN) -Ref [1].
Author Response: thank you for the suggestion. We have added your recent publication in the Introduction (see also page 4).
“The BORN birth registry has also been compared to another Canadian birth registry: CIHI-DAD (Canadian Institute for Health Information Discharge Abstract Database) [Darlings et al., 2024]. When lower agreement between variables was found, disagreement could often be explained by differences in coding or differences in definitions of the compared variables [bron: Darlings, 2024].”

We hope that our changes are satisfactory. We are looking forward to your evaluation.

Kind regards, on behalf of all authors,

Hilde Plomp
Competing Interests: No competing interests were disclosed. Close
Report a concern
Respond or Comment

COMMENTS ON THIS REPORT

Author Response 09 Jan 2025

Hilde Plomp, Midwifery Science, Amsterdam University Medical Centres, Amsterdam, The Netherlands

09 Jan 2025

Author Response

Dear Elizabeth K. Darling,

We would like to thank you for reading and reviewing the article, for the positive evaluation and compliments concerning our article, and for the additional ... Continue reading Dear Elizabeth K. Darling,

We would like to thank you for reading and reviewing the article, for the positive evaluation and compliments concerning our article, and for the additional thoughtful comments and suggestions. Below is a point-by-point reply to your comments. We indicated the start of your comments by using your initials (EKD) and our respective reply starts with HP (i.e. Hilde Plomp the first author answering on behalf of all the authors). Changed text passages, which are presented below, are indicated by using quotation marks, and for the convenience of the reviewer, we also refer to the respective pages and lines in the manuscript where these passages can be found.

Reviewer: Revising the titles of these tables might help to make their purpose more clear to the reader.
Author Response: We have changed the titles of the tables to make it more clear to the reader that these are example tables and what they contain. For example, the title of Table 1 now reads:
“Table 1. Example|Descriptive statistics of the variables to be studied based on Perined and the hospital records/CRFs.”

Reviewer: First, no details are provided regarding how the data sources will be linked. The authors cite a prior study that linked Perined and IRIS data, implicitly demonstrating feasibility, but do not explain what variable(s) will be used for linkage.”
Author Response: We have added some additional information about the data linkage procedure in the Methods section, as shown below (see also page 6 of the manuscript):
“Extracted data from the selected hospital records and CRFs were linked to those from the Perined records of the same mothers and neonates in order to compare these two types of records. Data managers linked the records by matching several demographic variables (e.g. date of birth of the mother and postal code). Variables chosen for analysis will be included in the planned study if they are available in both Perined and either the extracted hospital record data or the CRFs.”

Reviewer: Furthermore, under 'Ethical considerations', the authors state that the analyses will be done using an anonymous data source, without patient names and study ID, which again raises the question of how the data will be linked. If a previously linked data set is being used, this should be explained.
Author Response: We have added some extra information in the ‘Ethical considerations’ section. It is important to note that the datasets were not completely anonymous to the data managers involved in our study. However, researchers receive anonymized datasets for data-analysis purposes. Our additions to the ‘Ethical considerations’ section are shown below (see also page 14):
“Women gave permission to link their records, the linking was performed by data managers of the IRIS study with the support of Perined. After linking the data all identifiable patient information about patients was removed. Analyses for the planned study will be done with a completely anonymous data set, without any identifying variables (e.g. patient names, date of births and study ID).”

Reviewer: Second, the protocol for the study should include the algorithms that will be used to recode some variables to allow for comparison between two sources. This could be presented as the first table (or a supplementary table if it is very lengthy), and should ideally include the name of the variable in each source and the rules that will be applied to recode the comparator variables to align them with Perined variables when applicable. This level of detail is necessary to permit replication of the study.
Author Response: We cannot include the requested syntax in the current article as the stage of our planned work does not allow this. The syntax of the recoding will be published as a supplementary file included in the subsequent scientific article reporting our final results. Until we have completed the variable coding and data-analysis, we cannot guarantee full precision regarding the best recoding method for each variable (see also page 6):
“The categories of the variables will be created with data driven approaches. We expect that the data for these variables as derived from the respective source can be coded into the categories as shown in Tables 1 and 2, or are already available in this way. The syntax comprising the respective coding steps will be published, after publication of the final results as planned in the current study. In case a good comparison between certain variables is not possible, they will be removed and not included in the comparative analyses.”

Reviewer: Finally, it would be helpful for the reader if you could add another sentence or two to explain the difference between the old and new registration approaches for hospitals that are/were used for Perined. This additional contextual detail would help the reader understand the potential implications of the change in approach.
Author Response: We have added some additional information on this aspect in the Methods section (see also page 5):
“In 2015 and 2016 Perined started with a new registration approach for hospitals in the Netherlands. Some hospitals used the new approach, while others still used the old approach. The old and the new approach differ in the variables that are recorded, some variables have been added in the new approach, while others have been removed. Also, for some variables answer options have been changed. For the IRIS study, most of the data were provided according to the old approach. The data provided according to the new approach have been recoded by Perined according to the specifications of the old approach.”

Reviewer: Given the citation of two similar studies in the introduction, the authors may be interested in this recent publication that also examines data accuracy of the Ontario Birth Registry (BORN) -Ref [1].
Author Response: thank you for the suggestion. We have added your recent publication in the Introduction (see also page 4).
“The BORN birth registry has also been compared to another Canadian birth registry: CIHI-DAD (Canadian Institute for Health Information Discharge Abstract Database) [Darlings et al., 2024]. When lower agreement between variables was found, disagreement could often be explained by differences in coding or differences in definitions of the compared variables [bron: Darlings, 2024].”

We hope that our changes are satisfactory. We are looking forward to your evaluation.

Kind regards, on behalf of all authors,

Hilde Plomp
Dear Elizabeth K. Darling,

We would like to thank you for reading and reviewing the article, for the positive evaluation and compliments concerning our article, and for the additional thoughtful comments and suggestions. Below is a point-by-point reply to your comments. We indicated the start of your comments by using your initials (EKD) and our respective reply starts with HP (i.e. Hilde Plomp the first author answering on behalf of all the authors). Changed text passages, which are presented below, are indicated by using quotation marks, and for the convenience of the reviewer, we also refer to the respective pages and lines in the manuscript where these passages can be found.

Reviewer: Revising the titles of these tables might help to make their purpose more clear to the reader.
Author Response: We have changed the titles of the tables to make it more clear to the reader that these are example tables and what they contain. For example, the title of Table 1 now reads:
“Table 1. Example|Descriptive statistics of the variables to be studied based on Perined and the hospital records/CRFs.”

Reviewer: First, no details are provided regarding how the data sources will be linked. The authors cite a prior study that linked Perined and IRIS data, implicitly demonstrating feasibility, but do not explain what variable(s) will be used for linkage.”
Author Response: We have added some additional information about the data linkage procedure in the Methods section, as shown below (see also page 6 of the manuscript):
“Extracted data from the selected hospital records and CRFs were linked to those from the Perined records of the same mothers and neonates in order to compare these two types of records. Data managers linked the records by matching several demographic variables (e.g. date of birth of the mother and postal code). Variables chosen for analysis will be included in the planned study if they are available in both Perined and either the extracted hospital record data or the CRFs.”

Reviewer: Furthermore, under 'Ethical considerations', the authors state that the analyses will be done using an anonymous data source, without patient names and study ID, which again raises the question of how the data will be linked. If a previously linked data set is being used, this should be explained.
Author Response: We have added some extra information in the ‘Ethical considerations’ section. It is important to note that the datasets were not completely anonymous to the data managers involved in our study. However, researchers receive anonymized datasets for data-analysis purposes. Our additions to the ‘Ethical considerations’ section are shown below (see also page 14):
“Women gave permission to link their records, the linking was performed by data managers of the IRIS study with the support of Perined. After linking the data all identifiable patient information about patients was removed. Analyses for the planned study will be done with a completely anonymous data set, without any identifying variables (e.g. patient names, date of births and study ID).”

Reviewer: Second, the protocol for the study should include the algorithms that will be used to recode some variables to allow for comparison between two sources. This could be presented as the first table (or a supplementary table if it is very lengthy), and should ideally include the name of the variable in each source and the rules that will be applied to recode the comparator variables to align them with Perined variables when applicable. This level of detail is necessary to permit replication of the study.
Author Response: We cannot include the requested syntax in the current article as the stage of our planned work does not allow this. The syntax of the recoding will be published as a supplementary file included in the subsequent scientific article reporting our final results. Until we have completed the variable coding and data-analysis, we cannot guarantee full precision regarding the best recoding method for each variable (see also page 6):
“The categories of the variables will be created with data driven approaches. We expect that the data for these variables as derived from the respective source can be coded into the categories as shown in Tables 1 and 2, or are already available in this way. The syntax comprising the respective coding steps will be published, after publication of the final results as planned in the current study. In case a good comparison between certain variables is not possible, they will be removed and not included in the comparative analyses.”

Reviewer: Finally, it would be helpful for the reader if you could add another sentence or two to explain the difference between the old and new registration approaches for hospitals that are/were used for Perined. This additional contextual detail would help the reader understand the potential implications of the change in approach.
Author Response: We have added some additional information on this aspect in the Methods section (see also page 5):
“In 2015 and 2016 Perined started with a new registration approach for hospitals in the Netherlands. Some hospitals used the new approach, while others still used the old approach. The old and the new approach differ in the variables that are recorded, some variables have been added in the new approach, while others have been removed. Also, for some variables answer options have been changed. For the IRIS study, most of the data were provided according to the old approach. The data provided according to the new approach have been recoded by Perined according to the specifications of the old approach.”

Reviewer: Given the citation of two similar studies in the introduction, the authors may be interested in this recent publication that also examines data accuracy of the Ontario Birth Registry (BORN) -Ref [1].
Author Response: thank you for the suggestion. We have added your recent publication in the Introduction (see also page 4).
“The BORN birth registry has also been compared to another Canadian birth registry: CIHI-DAD (Canadian Institute for Health Information Discharge Abstract Database) [Darlings et al., 2024]. When lower agreement between variables was found, disagreement could often be explained by differences in coding or differences in definitions of the compared variables [bron: Darlings, 2024].”

We hope that our changes are satisfactory. We are looking forward to your evaluation.

Kind regards, on behalf of all authors,

Hilde Plomp
Competing Interests: No competing interests were disclosed. Close
Report a concern

Views

Reviewer Report 13 Jul 2024

Patricia Lee King, University of Chicago Pritzker School of Medicine, Chicago, USA; Medical Social Sciences, Northwestern University - Chicago (Ringgold ID: 205058), Chicago, Illinois, USA

Approved with Reservations

https://doi.org/10.5256/f1000research.164702.r295927

This article describes a research protocol for a planned evaluation of data quality of the Netherlands Perinatal Registry (Perined) comparing it to the IRIS study database. Intra-class correlation and Kappa will be calculated to determine accuracy. This research protocol will assess the accuracy of Perined which is widely used for research and public health decision making by comparing it to clinical data.

This is a detailed description of a study protocol of great importance to public health and research 9- the clinical accuracy of a large data set used for national decision making. It could be strengthened by clarifying in the article that this is a planned study/ study protocol being described. The structure of the article leaves the reader looking for results until they reach the conclusion section. The reference to "study protocol" article category at the title is probably not enough for clarity. The Tables, instead of example, could be labeled data collection forms/templates for further clarity.

Is the rationale for, and objectives of, the study clearly described?

Partly
Is the study design appropriate for the research question?

Partly
Are sufficient details of the methods provided to allow replication by others?

Yes
Are the datasets clearly presented in a useable and accessible format?

Not applicable

Competing Interests: No competing interests were disclosed.

Reviewer Expertise: quality improvement, perinatal health

CITE

Report a concern

Author Response 09 Jan 2025

Hilde Plomp, Midwifery Science, Amsterdam University Medical Centres, Amsterdam, The Netherlands

09 Jan 2025

Author Response

Dear Patricia Lee King,

We would like to thank you for the positive evaluation and compliments concerning our article and for your thoughtful comments. We agree with your suggestion ... Continue reading Dear Patricia Lee King,

We would like to thank you for the positive evaluation and compliments concerning our article and for your thoughtful comments. We agree with your suggestion to better clarify that the current paper concerns a study protocol, this was made clear in the title, but possibly insufficiently in the rest of the article. Therefore, we now explicitly state in several relevant places in the text and in the tables that the current paper concerns a study protocol (see the manuscript). Moreover, we have made some additional changes to the text by for example by using the term “planned study” to indicate that the current work only concerns a study protocol.

We hope that our changes are satisfactory. We look forward to your evaluation.

Kind regards, on behalf of all authors,

Hilde Plomp
Dear Patricia Lee King,

We would like to thank you for the positive evaluation and compliments concerning our article and for your thoughtful comments. We agree with your suggestion to better clarify that the current paper concerns a study protocol, this was made clear in the title, but possibly insufficiently in the rest of the article. Therefore, we now explicitly state in several relevant places in the text and in the tables that the current paper concerns a study protocol (see the manuscript). Moreover, we have made some additional changes to the text by for example by using the term “planned study” to indicate that the current work only concerns a study protocol.

We hope that our changes are satisfactory. We look forward to your evaluation.

Kind regards, on behalf of all authors,

Hilde Plomp
Competing Interests: No competing interests were disclosed. Close
Report a concern
Respond or Comment

COMMENTS ON THIS REPORT

Author Response 09 Jan 2025

Hilde Plomp, Midwifery Science, Amsterdam University Medical Centres, Amsterdam, The Netherlands

09 Jan 2025

Author Response

Dear Patricia Lee King,

We would like to thank you for the positive evaluation and compliments concerning our article and for your thoughtful comments. We agree with your suggestion ... Continue reading Dear Patricia Lee King,

We would like to thank you for the positive evaluation and compliments concerning our article and for your thoughtful comments. We agree with your suggestion to better clarify that the current paper concerns a study protocol, this was made clear in the title, but possibly insufficiently in the rest of the article. Therefore, we now explicitly state in several relevant places in the text and in the tables that the current paper concerns a study protocol (see the manuscript). Moreover, we have made some additional changes to the text by for example by using the term “planned study” to indicate that the current work only concerns a study protocol.

We hope that our changes are satisfactory. We look forward to your evaluation.

Kind regards, on behalf of all authors,

Hilde Plomp
Dear Patricia Lee King,

We would like to thank you for the positive evaluation and compliments concerning our article and for your thoughtful comments. We agree with your suggestion to better clarify that the current paper concerns a study protocol, this was made clear in the title, but possibly insufficiently in the rest of the article. Therefore, we now explicitly state in several relevant places in the text and in the tables that the current paper concerns a study protocol (see the manuscript). Moreover, we have made some additional changes to the text by for example by using the term “planned study” to indicate that the current work only concerns a study protocol.

We hope that our changes are satisfactory. We look forward to your evaluation.

Kind regards, on behalf of all authors,

Hilde Plomp
Competing Interests: No competing interests were disclosed. Close
Report a concern

Comments on this article Comments (0)

Version 2

VERSION 2 PUBLISHED 24 Jun 2024

Open Peer Review

Reviewer Status

Reviewer Reports

	Invited Reviewers
	1	2
Version 2 (revision) 09 Jan 25	read	read
Version 1 24 Jun 24	read	read

Patricia Lee King, University of Chicago Pritzker School of Medicine, Chicago, USA; Northwestern University - Chicago (Ringgold ID: 205058), Chicago, USA
Elizabeth Kathleen Darling, McMaster University, Hamilton, Canada

Comments on this article

All Comments(0)

Add a comment

Browse by related subjects

Back to all reports

Reviewer Report

1 Views

12 Feb 2025 | for Version 2

Elizabeth Kathleen Darling, McMaster University, Hamilton, Ontario, Canada

1 Views Cite this report Responses(0)

Approved

Thank you for your thoughtful and detailed responses to my previous comments. You have addressed all of my suggestions in a satisfactory manner.

Competing Interests

No competing interests were disclosed.

I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard.

Respond to this report

Responses (0)

Back to all reports

Reviewer Report

1 Views

03 Feb 2025 | for Version 2

Patricia Lee King, University of Chicago Pritzker School of Medicine, Chicago, USA; Medical Social Sciences, Northwestern University - Chicago (Ringgold ID: 205058), Chicago, Illinois, USA

1 Views Cite this report Responses(0)

Approved

Thank you to the authors for the edits based on reviewer feedback. They were responsive to the requests. My only remaining comment is that I do still think the use of the word study in the title and the paper impacts clarity that this is a protocol for a study that will happen and confuses the reader on what is a plan versus what is completed. For example, with the following edits to the title I'm more clear on what is a compelted study and what is work forthcoming:
A research protocol to evaluate data quality in the Netherlands Perinatal Registry (Perined): A protocol for a data comparison study using hospital records from the IUGR Risk Selection (IRIS) study

Competing Interests

No competing interests were disclosed.

Reviewer Expertise

quality improvement, perinatal health

I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard.

Respond to this report

Responses (0)

Back to all reports

Reviewer Report

16 Views

16 Jul 2024 | for Version 1

Elizabeth Kathleen Darling, McMaster University, Hamilton, Ontario, Canada

16 Views Cite this report Responses(1)

Approved With Reservations

Is the rationale for, and objectives of, the study clearly described?

Yes
Is the study design appropriate for the research question?

Yes
Are sufficient details of the methods provided to allow replication by others?

No
Are the datasets clearly presented in a useable and accessible format?

Not applicable

References

Competing Interests

No competing interests were disclosed.

Reviewer Expertise

midwifery services, large database research, access to care, health equity

Respond to this report

Responses (1)

Author Response

09 Jan 2025

Hilde Plomp, Midwifery Science, Amsterdam University Medical Centres, Amsterdam, The Netherlands

Dear Elizabeth K. Darling,

We would like to thank you for reading and reviewing the article, for the positive evaluation and compliments concerning our article, and for the additional thoughtful comments and suggestions. Below is a point-by-point reply to your comments. We indicated the start of your comments by using your initials (EKD) and our respective reply starts with HP (i.e. Hilde Plomp the first author answering on behalf of all the authors). Changed text passages, which are presented below, are indicated by using quotation marks, and for the convenience of the reviewer, we also refer to the respective pages and lines in the manuscript where these passages can be found.

Reviewer: Revising the titles of these tables might help to make their purpose more clear to the reader.
Author Response: We have changed the titles of the tables to make it more clear to the reader that these are example tables and what they contain. For example, the title of Table 1 now reads:
“Table 1. Example|Descriptive statistics of the variables to be studied based on Perined and the hospital records/CRFs.”

Reviewer: First, no details are provided regarding how the data sources will be linked. The authors cite a prior study that linked Perined and IRIS data, implicitly demonstrating feasibility, but do not explain what variable(s) will be used for linkage.”
Author Response: We have added some additional information about the data linkage procedure in the Methods section, as shown below (see also page 6 of the manuscript):
“Extracted data from the selected hospital records and CRFs were linked to those from the Perined records of the same mothers and neonates in order to compare these two types of records. Data managers linked the records by matching several demographic variables (e.g. date of birth of the mother and postal code). Variables chosen for analysis will be included in the planned study if they are available in both Perined and either the extracted hospital record data or the CRFs.”

Reviewer: Furthermore, under 'Ethical considerations', the authors state that the analyses will be done using an anonymous data source, without patient names and study ID, which again raises the question of how the data will be linked. If a previously linked data set is being used, this should be explained.
Author Response: We have added some extra information in the ‘Ethical considerations’ section. It is important to note that the datasets were not completely anonymous to the data managers involved in our study. However, researchers receive anonymized datasets for data-analysis purposes. Our additions to the ‘Ethical considerations’ section are shown below (see also page 14):
“Women gave permission to link their records, the linking was performed by data managers of the IRIS study with the support of Perined. After linking the data all identifiable patient information about patients was removed. Analyses for the planned study will be done with a completely anonymous data set, without any identifying variables (e.g. patient names, date of births and study ID).”

Reviewer: Second, the protocol for the study should include the algorithms that will be used to recode some variables to allow for comparison between two sources. This could be presented as the first table (or a supplementary table if it is very lengthy), and should ideally include the name of the variable in each source and the rules that will be applied to recode the comparator variables to align them with Perined variables when applicable. This level of detail is necessary to permit replication of the study.
Author Response: We cannot include the requested syntax in the current article as the stage of our planned work does not allow this. The syntax of the recoding will be published as a supplementary file included in the subsequent scientific article reporting our final results. Until we have completed the variable coding and data-analysis, we cannot guarantee full precision regarding the best recoding method for each variable (see also page 6):
“The categories of the variables will be created with data driven approaches. We expect that the data for these variables as derived from the respective source can be coded into the categories as shown in Tables 1 and 2, or are already available in this way. The syntax comprising the respective coding steps will be published, after publication of the final results as planned in the current study. In case a good comparison between certain variables is not possible, they will be removed and not included in the comparative analyses.”

Reviewer: Finally, it would be helpful for the reader if you could add another sentence or two to explain the difference between the old and new registration approaches for hospitals that are/were used for Perined. This additional contextual detail would help the reader understand the potential implications of the change in approach.
Author Response: We have added some additional information on this aspect in the Methods section (see also page 5):
“In 2015 and 2016 Perined started with a new registration approach for hospitals in the Netherlands. Some hospitals used the new approach, while others still used the old approach. The old and the new approach differ in the variables that are recorded, some variables have been added in the new approach, while others have been removed. Also, for some variables answer options have been changed. For the IRIS study, most of the data were provided according to the old approach. The data provided according to the new approach have been recoded by Perined according to the specifications of the old approach.”

Reviewer: Given the citation of two similar studies in the introduction, the authors may be interested in this recent publication that also examines data accuracy of the Ontario Birth Registry (BORN) -Ref [1].
Author Response: thank you for the suggestion. We have added your recent publication in the Introduction (see also page 4).
“The BORN birth registry has also been compared to another Canadian birth registry: CIHI-DAD (Canadian Institute for Health Information Discharge Abstract Database) [Darlings et al., 2024]. When lower agreement between variables was found, disagreement could often be explained by differences in coding or differences in definitions of the compared variables [bron: Darlings, 2024].”

We hope that our changes are satisfactory. We are looking forward to your evaluation.

Kind regards, on behalf of all authors,

Hilde Plomp

View more View less

Competing Interests

No competing interests were disclosed.

Back to all reports

Reviewer Report

19 Views

13 Jul 2024 | for Version 1

Patricia Lee King, University of Chicago Pritzker School of Medicine, Chicago, USA; Medical Social Sciences, Northwestern University - Chicago (Ringgold ID: 205058), Chicago, Illinois, USA

19 Views Cite this report Responses(1)

Approved With Reservations

Is the rationale for, and objectives of, the study clearly described?

Partly
Is the study design appropriate for the research question?

Partly
Are sufficient details of the methods provided to allow replication by others?

Yes
Are the datasets clearly presented in a useable and accessible format?

Not applicable

Competing Interests

No competing interests were disclosed.

Reviewer Expertise

quality improvement, perinatal health

Respond to this report

Responses (1)

Author Response

09 Jan 2025

Hilde Plomp, Midwifery Science, Amsterdam University Medical Centres, Amsterdam, The Netherlands

Dear Patricia Lee King,

We would like to thank you for the positive evaluation and compliments concerning our article and for your thoughtful comments. We agree with your suggestion to better clarify that the current paper concerns a study protocol, this was made clear in the title, but possibly insufficiently in the rest of the article. Therefore, we now explicitly state in several relevant places in the text and in the tables that the current paper concerns a study protocol (see the manuscript). Moreover, we have made some additional changes to the text by for example by using the term “planned study” to indicate that the current work only concerns a study protocol.

We hope that our changes are satisfactory. We look forward to your evaluation.

Kind regards, on behalf of all authors,

Hilde Plomp

View more View less

Competing Interests

No competing interests were disclosed.

Alongside their report, reviewers assign a status to the article:

Approved - the paper is scientifically sound in its current form and only minor, if any, improvements are suggested

Approved with reservations - A number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.

Not approved - fundamental flaws in the paper seriously undermine the findings and conclusions

[1] 1. Phillips DE, Adair T, Lopez AD: How useful are registered birth statistics for health and social policy? A global systematic assessment of the availability and quality of birth registration data. Popul. Health Metrics. 2018; 16(1): 21. PubMed Abstract | Publisher Full Text | Free Full Text

[2] 2. Project Euro-Peristat: European Perinatal Health Report – Core indicators of the and care of pregnant women and babies in Europe from 2015 to 2019. Paris: Euro-Peristat; 2022.

[3] 3. Delnord M, Szamotulska K, Hindori-Mohangoo AD, et al.: Linking databases on perinatal health: a review of the literature and current practices in Europe. Eur. J. Pub. Health. 2016; 26(3): 422–430. PubMed Abstract | Publisher Full Text | Free Full Text

[4] 4. Perined: Perinatale zorg in Nederland anno 2020 [Perinatal care in the Netherlands in 2020]. Utrecht: Perined; 2021.

[5] 5. Devlieger RME, Goemaes R, Cammu H: Perinatale activiteiten in Vlaanderen 2017 [Perinatal activities in Flanders 2017]. Brussel: SPE; 2018.

[6] 6. Draper ESGI, Smith LK, Kurinczuk JJ, et al.: MBRRACE-UK Perinatal Mortality Surveillance Report UK Perinatal Deaths for Births from January to December 2017. Leicester: MBRRACE-UK collaboration; 2019.

[7] 7. Miao Q, Fell DB, Dunn S, et al.: Agreement assessment of key maternal and newborn data elements between birth registry and Clinical Administrative Hospital Databases in Ontario, Canada. Arch. Gynecol. Obstet. 2019; 300(1): 135–143. PubMed Abstract | Publisher Full Text

[8] 8. Dunn SLA, Sprague AE, Fell DB, et al.: Data accuracy in the Ontario birth Registry: a chart re-abstraction study. BMC Health Serv. Res. 2019; 19: 1001–1011. PubMed Abstract | Publisher Full Text | Free Full Text

[9] 9. Darling EK, Marquez O, Park AL, et al.: Defining a low-risk birth cohort: a cohort study comparing two perinatal data sets in Ontario, Canada. Int. J. Popul. Data Sci. 2024; 9(1): 2364. PubMed Abstract | Publisher Full Text | Free Full Text

[10] 10. Gissler M, Shelley J: Quality of data on subsequent events in a routine Medical Birth Register. Med. Inform. Internet Med. 2002; 27(1): 33–38. PubMed Abstract | Publisher Full Text

[11] 11. Ballantine KR, Hanna S, Macfarlane S, et al.: Childhood cancer registration in New Zealand: A registry collaboration to assess and improve data quality. Cancer Epidemiol. 2018; 55: 104–109. PubMed Abstract | Publisher Full Text

[12] 12. Linder G, Lindblad M, Djerf P, et al.: Validation of data quality in the Swedish National Register for Oesophageal and Gastric Cancer. Br. J. Surg. 2016; 103(10): 1326–1335. PubMed Abstract | Publisher Full Text

[13] 13. Londero SC, Mathiesen JS, Krogdahl A, et al.: Completeness and validity in a national clinical thyroid cancer database: DATHYRCA. Cancer Epidemiol. 2014; 38(5): 633–637. PubMed Abstract | Publisher Full Text

[14] 14. Ostgard LS, Norgaard JM, Severinsen MT, et al.: Data quality in the Danish National Acute Leukemia Registry: a hematological data resource. Clin. Epidemiol. 2013; 5: 335–344. PubMed Abstract | Publisher Full Text | Free Full Text

[15] 15. Lofgren L, Eloranta S, Krawiec K, et al.: Validation of data quality in the Swedish National Register for Breast Cancer. BMC Public Health. 2019; 19(1): 495. PubMed Abstract | Publisher Full Text | Free Full Text

[16] 16. Perined: Over Perined [About Perined]. Website Perined. Accessed January 23, 2023. Reference Source

[17] 17. Mannien J, Klomp T, Wiegers T, et al.: Evaluation of primary care midwifery in The Netherlands: design and rationale of a dynamic cohort study (DELIVER). BMC Health Serv. Res. 2012; 12: 69. PubMed Abstract | Publisher Full Text | Free Full Text

[18] 18. Van Eijsden M, Vrijkotte TG, Gemke RJ, et al.: Cohort profile: the Amsterdam Born Children and their Development (ABCD) study. Int. J. Epidemiol. 2011; 40(5): 1176–1186. PubMed Abstract | Publisher Full Text

[19] 19. Henrichs J, Verfaille V, Jellema P, et al.: Effectiveness of routine third trimester ultrasonography to reduce adverse perinatal outcomes in low risk pregnancy (the IRIS study): nationwide, pragmatic, multicentre, stepped wedge cluster randomised trial. BMJ. 2019; 367: l5517. PubMed Abstract | Publisher Full Text | Free Full Text

[20] 20. Hoftiezer L, Hof MPH, Dijs-Elsinga J, et al.: From population reference to national standard: new and improved birthweight charts. Am. J. Obstet. Gynecol. 2019; 220(4): 383.e1–383.e17. PubMed Abstract | Publisher Full Text

[21] 21. Henrichs J, Verfaille V, Viester L, et al.: Effectiveness and cost-effectiveness of routine third trimester ultrasound screening for intrauterine growth restriction: study protocol of a nationwide stepped wedge cluster-randomized trial in The Netherlands (The IRIS Study). BMC Pregnancy Childbirth. 2016; 16(1): 310. Publisher Full Text

[22] 22. Perined: Perinatale Zorg in Nederland 2016 [Perinatale Care in the Netherlands 2016]. Utrecht: Perined; 2018.

[23] 23. Perined: Registratie [Registration]. Website Perined. Accessed January 23, 2023. Reference Source

[24] 24. Perined: Methodologie Perinatologie Data: Van ontvangst tot gebruiksklare data voor onderzoek, benchmarking en rapportage [Methodology Perinatology Data: From receipt to ready-to-use data for research, benchmarking and reporting]. Utrecht: 2020.

[25] 25. Henrichs J, De Jonge A, Westerneng M, et al.: Cost-Effectiveness of Routine Third Trimester Ultrasound Screening for Fetal Growth Restriction Compared to Care as Usual in Low-Risk Pregnancies: A Pragmatic Nationwide Stepped-Wedge Cluster-Randomized Trial in The Netherlands (the IRIS Study). Int. J. Environ. Res. Public Health. 2022; 19(3312). PubMed Abstract | Publisher Full Text | Free Full Text

[26] 26. Byrt T, Bishop J, Carlin JB: Bias, prevalence and kappa. J. Clin. Epidemiol. 1993; 46(5): 423–429. Publisher Full Text

[27] 27. Vannest KJ, Parker RI, Gonen O, et al.: Single Case Research: web based calculators for SCR analysis. (Version 2.0) [Web-based application]. Accessed February 7, 2024. Reference Source

[28] 28. Twisk J: Inleiding in de toegepaste biostatistiek [Introduction in applied biostatistics]. Houten: Bohn Stafleu van Loghum; 2017.

[29] 29. Landis JR, Koch GG: The measurement of observer agreement for categorical data. Biometrics. 1977; 33(1): 159–174. Publisher Full Text

[30] 30. Viera AJ, Garrett JM: Understanding interobserver agreement: the kappa statistic. Fam. Med. 2005; 37(5): 360–363. PubMed Abstract

[31] 31. Feinstein AR, Cicchetti DV: High agreement but low kappa: I. The problems of two paradoxes. J. Clin. Epidemiol. 1990; 43(6): 543–549. PubMed Abstract | Publisher Full Text

[32] 32. Koo TK, Li MY: A Guideline of Selecting and Reporting Intraclass Correlation Coefficients for Reliability Research. J. Chiropr. Med. 2016; 15(2): 155–163. PubMed Abstract | Publisher Full Text | Free Full Text

[33] 33. Becking EC, Scheffer PG, Henrichs J, et al.: Fetal fraction of cell-free DNA in noninvasive prenatal testing and adverse pregnancy outcomes: a nationwide retrospective cohort study of 56,110 pregnant women. Am. J. Obstet. Gynecol. 2023. PubMed Abstract | Publisher Full Text

[34] 34. The Helsinki Declaration of the World Medical Association (WMA): Ethical principles of medical research involving human subjects. Pol Merkur Lekarski. 2014; 36: 301. Publisher Full Text

[35] 35. RStudio Team: RStudio: Integrated Development for R. Boston, MA: RStudio; 2020.

[36] 36. IBM Corp: IBM SPSS Statistics for Windows, Version 28.0. Armonk, NY: IBM Corp; 2021.

[37] 37. Pfaff B, Darrington J, Stover J, et al.: GNU PSPP Statistical Analysis Software: Release 0.9. Boston, MA: Free Software Foundation; 2007.

Research protocol - Evaluating data quality in the Netherlands Perinatal Registry (Perined): A data comparison study using hospital records from the IUGR Risk Selection (IRIS) study

Abstract

Background

Methods

Discussion

Keywords

Revised Amendments from Version 1

Introduction

Methods

Design

Perined data

Figure 1. Perined links datasets and creates one record for all neonates and their mothers.24

Data collection and selection

Figure 2. Data sources of the planned study.

Table 1. Example|Descriptive statistics of the variables to be studied based on Perined and the hospital records/CRFs.

Table 2. Example|Degree of completeness and the agreement between Perined and hospital records/CRFs to be studied.

Statistical analyses

Conclusion and discussion

Ethical considerations

Consent to participate

Author contributions

Supplementary material

Data availability

Software availability

Acknowledgements

References

Comments on this article Comments (0)

Open Peer Review

Comments on this article Comments (0)

Open Peer Review

Reviewer Status

Reviewer Reports

Comments on this article

Browse by related subjects

Competing Interests Policy

Stay Updated

Figure 1. Perined links datasets and creates one record for all neonates and their mothers.²⁴