Pre-hospital delay in Vietnamese patients hospitalized with a first acute myocardial infarction: A short report

Ethical statement

This study was approved by the Institutional Review Board at the Institute of Population, Health and Development. A waiver of patient consent was approved by the Institutional Review Board. All patients’ personal information was de-identified before analysis.

Study setting

This study was conducted at the Vietnam National Heart Institute (VNHI)^3,4. This hospital is a 250 bed tertiary care medical center in Hanoi (2009 census = 6.5 million) which manages the majority of Hanoi residents hospitalized with acute coronary disease and other NCDs.

Case ascertainment

Computerized printouts of discharged patients from the VNHI in 2010 with possible AMI were collected and International Classification of Disease codes for possible AMI (I20–I25) were reviewed. Two cardiologists validated these cases utilizing predefined criteria developed by the World Health Organization. They include having a suggestive medical history, serum enzyme increasing above the hospital’s reference range, and serial electrocardiographic (ECG) findings during hospitalization consistent with the presence of AMI; at least two of these three criteria needed to be present for an AMI to have occurred⁵. There were 315 Hanoi residents, who satisfied the criteria for AMI were identified. However, 13 (4%) patients with history of AMI and 199 (66%) patients transferred from other facilities were excluded from the present analysis. Patients were then further divided into groups of STEMI and non-STEMI, utilizing standard classification techniques⁶. All ECGs of potential AMI patients were reviewed by a physician under the supervision of a senior cardiologist. Due to reviewed ECGs’ inadequate quality, nine patients with AMI could not be classified as STEMI or non-STEMI.

Data collection

For all AMI-validated cases through the aforementioned process of independent adjudication utilizing eligibility criteria, information on social demographics, clinical, medical management and hospital discharge status were collected from medical record reviews and recorded onto a standardized case report form by trained study physicians.

Pre-hospital delay is the primary outcome investigated by this study. The delay time was defined as the time interval between the onset of symptoms suggestive of AMI and the patient’s arrival at medical centers^7,8. Our trained physicians reviewed any information they could seek in medical records, which described the duration of pre-hospital delay from emergency personnel, nurses, and physicians notes. Information on pre-hospital delay was collected in minutes (as a continuous variable) then was further categorized according to cut-points of 6 and 12 hours, which were commonly applied in prior studies, and based on the data distribution Patients, in whom no exact time of symptom onset was reported, and only had a delay time interval recorded in their medical record (e.g. <6, 6-<12, and ≥12 hours), were incorporated into our analysis of delay time when it was constructed as a categorical outcome. We restricted our patient population to those with information available on prehospital delay (either exact time noted or time interval) in their hospital medical records.

Data analysis

The overall mean (standard deviation-SD) and median (inter quartile range-IQR) duration of pre-hospital delay was calculated according to standard methods. Data were shown as percentages for categorical variables and compared between patients who delayed ≥ 6 hours, and patients who sought medical care earlier, using chi-square or Fisher exact tests; medians (inter quartile range-IQR) for continuous variables were calculated and compared using the Wilcoxon sum rank tests. Since the sample size in this study was small, no regression analysis was performed. All analyses were performed using STATA 11.0 (StataCorp. TX).

Extent of pre-hospital delay

Information on pre-hospital delay was collected in minutes (as a continuous variable) for 63% (n=65) of patients. In 37% of patients (n=38), an exact time of symptom onset was not reported, and only a delay time interval was recorded in patient’s medical record (e.g. <6, 6-<12, and ≥12 hours); these patients were included in the analysis of delay time when it was constructed as a categorical outcome.

Analysis of prehospital delay as a continuous outcome. Within the group of 65 patients reporting an exact time of symptom onset, the overall mean and median durations of pre-hospital delay were 14.9 hours (SD: 23 hours) and 4.8 hours (IQR: 3–10 hours), respectively (Figure 1). Among patients with STEMI (n=42), mean and median delay times were 16.4 hours (SD: 26 hours) and 4.8 hours (IQR: 3.0–6.2 hours), respectively. Among patients with non-STEMI (n=19), mean and median delay times were 14.0 hours (SD: 19.2 hours) and 8.3 hours (IQR: 2.5 -12.3 hours), respectively.

Figure 1. Mean and Median of Pre-hospital Delay in Patients Hospitalized with an initial Acute Myocardial Infarction.

Analysis of prehospital delay as a categorical outcome. When combining data from these patients (n=65), with more non-specific data from patients who reported duration of delay in a time interval (n=38), the proportion of patients who delayed <6 hours, 6-<12 hours, and ≥12 hours were 45%, 13%, and 42%, respectively (Figure 2). Among patients with STEMI (n=65), these proportions were 51%, 12% and 37%, respectively. Among patients with non-STEMI (n=29), these proportions were 31%, 17% and 52%, respectively.

Figure 2. Distribution of Duration of Pre-hospital Delay in Patients Hospitalized with an initial Acute Myocardial Infarction.

Patient characteristics associated with prolonged pre-Hospital delay

Patients who had a delay time over 6 hours were more likely to be female compared to their counterparts, who had a delay time of less than 6 hours (Table 1).

Study limitations

There are several imitations in this study. First, our data were collected from only one hospital, the VNHI in Hanoi, which limits the generalizability of our results to patients hospitalized in other medical centers across Hanoi. Some of observed patient characteristics may not reflect the general population of patients hospitalized with AMI in Hanoi; the distribution of non-STEMI appeared to be low in this study sample, though the frequency of various manifestations of AMI in Vietnamese adults is unknown. Second, due to our small sample size, we failed to identify potential factors that might relate to the prolonged delay time, information which is important for designing and implementing various educational intervention approaches to improve patient’s care seeking behavior. The subgroup analysis according to type of AMI, namely by STEMI and non-STEMI, was subject to potential bias in interpretation due to the small study sample size. Third, we were only able to calculate total duration of pre-hospital delay and we were unable to separately calculate patient delay time and extent of transportation delay since the time when patients made a treatment-seeking decision was not recorded in hospital medical records. In Vietnam, patients with signs and symptoms suggestive of AMI normally call for an ambulance or reach the hospital by themselves, and extremely few patients have a primary care doctor to consult with. In terms of transportation time, this study included only patients who reside in Hanoi. Forth, due to the retrospective design of the study, we were unable to obtain systematically collected and/or recorded information about the time of onset of symptoms suggestive of AMI in hospital medical records. This information was abstracted from notes written by different physicians, nurses, and emergency personnel which may have been collected and recorded in a non-standardized manner. Data on duration of prehospital delay were missing in approximately two thirds of Hanoi residents hospitalized with independently validated AMI, which was mainly due to missing data on the time of onset of acute symptoms suggestive of AMI, and not emergency department arrival time. Finally, we were unable to obtain information on additional patient-associated characteristics (e.g., socioeconomic status [eg., Education, occupation] and psychological factors [eg., reluctance to call family members for help or having no one around when experiencing symptoms to ask for help]) which may have impacted on patient’s medical care seeking behavior.

Our data show that patients admitted for initial AMI in Hanoi, Vietnam experienced a significant delay time in seeking medical attention after the onset of AMI suggestive symptoms. Additional full-scale data of hospitalized patients with AMI across Hanoi will be needed in order to confirm these preliminary descriptive findings. Public educational programs for the general population, specifically targeting towards women, is greatly needed to incentivize patients to seek medical assistance immediately to ensure maximum effectiveness of treatments. Further research is required to understand potential factors associated with the prolonged delay in Vietnamese patients hospitalized with AMI and the most effective ways to encourage patients to seek medical care in a timely manner.