Review protocol presenting search strategy was established without a registration number.

The Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) checklist was used to guide the present study. ^{21 , 63} Two researchers (LD and JBL) independently searched for articles in PubMed (RRID:SCR_004846), PEDro and EMBASE (RRID:SCR_001650) by using MeSH keywords. The search strategy phrase for PubMed database was (“smartphone application” OR “smartphone app” OR “smartphone” OR “telerehabilitation” OR “telemedicine” OR “mhealth” OR “ehealth”) AND (“low back pain” OR “back pain”). Corresponding research was realized in PEDro and EMBASE by using the same keywords. Only English language and time frame filters were used for our research. The articles were screened and assessed for eligibility regarding the objective of the study. Articles published between January 1st 2005 and May 3rd 2021 were considered. This time frame includes the emergence of the new generation of mobile phones.

Only studies of smartphone use in LBP adult patient healthcare were included. We excluded studies concerning telemedicine or telerehabilitation but without use of a smartphone. Similarly, we excluded studies in which the smartphone use involved healthy individuals for preventing LBP or promoting physical activity or a healthy lifestyle in healthy individuals. To broaden our review, all qualitative and quantitative studies were accepted, including randomised controlled trials (RCTs) and non-randomized studies, cohort studies, case-control studies, systematic reviews, reliability and validity studies, and study protocols. In cases where both the protocol and the completed study article were published, we included both to ensure comprehensive content analysis of the literature on the subject. Only studies accepted for publication, written in English were considered.

Study inclusion was discussed to reach agreement or in cases where a consensus could not be reached, we consulted with a third researcher (EC). Then studies were classified according to the purpose of the smartphone app. The type of study design was also used to classify studies in terms of its objective: effectiveness of the smartphone use, reliability of the smartphone measures, or other type of evaluation and data collection.

Search strategy was prepared and validated with the participation of all authors of the present review. The same search strategy was realized by two researchers (LD and JBL) independently. A third researcher (EC) participated in the synthesis data charting. Then, additional articles were identified throughout citation matching.

We extracted data concerning smartphone app utilisation. We focused on the purpose of smartphone app utilisation, study design, date of publication, number of participants who completed the study, main outcomes, results and authors conclusions.

The quality of selected RCTs was estimated by using the PEDro scale. ²²

We categorized included studies by the purpose of smartphone app utilisation. The synthesis of the included studies was also done by study design, both were presented in a narrative format and in a synthesis table.

A total of 14 studies used smartphone apps as an evaluation tool including pain evaluation, ^{23 – 26} ROM measurement, ^{16 , 27 – 30} 6-min walk test ³¹ and clinical or functional evaluation. ^{17 , 32 – 34} Considering pain evaluation, PainMAP showed excellent intra- and inter-rater reliability and good validity for quantifying the number of pain sites and pain area. ²³ This application allows patients to mark their pain on digital body charts and automatically quantifies such self-assessment. Furthermore, another smartphone app could effectively evaluate pain changes after transcutaneous electrical nerve stimulation. ²⁵ However, Ross et al., reported that patients with chronic pain who appeared to manage their pain better were less likely to report benefits of a smartphone pain app designed for daily pain management and evaluation. ²⁶

Smartphone apps were used to measure lumbar spine ROM and lordosis but showed insufficient reliability and validity as compared with a gravity-based inclinometer. ²⁷ By contrast, de Brito Macedo et al., reported good concurrent validity and intra-rater reliability of smartphone thoracolumbar ROM measurements versus a digital inclinometer. ²⁹ Lumbar ROM measurement with a smartphone detected a significant difference due to manual therapy and exercise intervention in another study. ²⁸ Similar findings were further summarized in systematic review showing contradictory results of psychometric properties of the lumbar spine ROM and lordosis measurements with a smartphone app. ³⁰

Other examples of smartphone use included clinical and functional evaluation. Peterson et al., suggested that a modified treatment-based classification system for subgrouping patients with LBP could direct treatment in telerehabilitation settings (smartphone app) because the overall rate of percentage agreement with face-to-face assessments was between 48.9% and 59.6%. ¹⁷ The McKenzie classification and therapy applied with a smartphone app showed comparable clinical outcomes with the traditional clinic-based McKenzie therapy. ³² The 6-min walk test performed with a smartphone app and using a GPS system could be reliable but needed to be performed avoiding indoors and city environments with high buildings and rectangular walking curses. ³¹ However, in a scoping review of outcomes with smartphone apps used for LBP management, authors showed the problem of low quality of the outcome measures to monitor the treatment effect. Indeed, only a few smartphone apps offered to monitor their effectiveness in the management of LBP. ³³ Other examples of smartphone use for clinical evaluation is the use of the Backache Disability Index. Such evaluation includes rating five trunk movements in the erect position and scoring morning back stiffness. Performed remotely with a smartphone app, the index showed good reliability and validity. ³⁴

A total of 12 studies presented smartphone apps as a tool for providing rehabilitation interventions at a distance (telerehabilitation). ^{17 , 32 , 35 – 44} Despite different designs and aims, most of the studies evaluated the effectiveness of the smartphone app use in telerehabilitation on pain and disability. Dario et al., performed a systematic review with meta-analysis to evaluate the effectiveness of the intervention based on any form of telerehabilitation, phone calls, emailing, and web-based chats but also smartphone apps. On the basis of data from 11 included studies, the authors concluded moderate evidence that current telerehabilitation interventions are not more effective than minimal interventions on pain and disability outcomes. Their study also revealed that the effectiveness of such interventions remains understudied. ⁴⁰ Indeed, only three RCTs evaluated the effectiveness of the smartphone apps use to provide telerehabilitation in LBP patients. Hou et al., evaluated a system of telerehabilitation based on a smartphone app interface for patients after lumbar surgery for LBP and a web-based interface for physicians providing and surveying rehabilitation and communicating with the patient. The authors demonstrated that the intervention was more effective than usual care regarding disability and pain status (Ostwestry Disability Index, pain visual analogue scale) at 24-month follow-up. ³⁸ Mbada et al., in their RCT, compared telerehabilitation-based McKenzie therapy versus the same treatment but provided in the traditional face-to-face setting. The smartphone app interface was used to introduce the treatment and was supported by phone calls and text messages in the experimental group. In this study, clinical improvement was noted at 4- and 8-week follow-ups, but no difference was found between the two groups in pain intensity, back muscle endurance and activity limitation. Thus, the authors concluded that smartphone app-based McKenzie telerehabilitation can be successfully used especially to bridge the gap in the non-availability of clinic-based therapy. ³² Toelle et al., investigated the effect of a multidisciplinary smartphone app ( Kaia App) on pain intensity at 12-week follow-ups in an RCT design. The experimental group received multidisciplinary self-management treatment based on current guidelines and supported by a chat with a healthcare professional via the app. The experimental group showed significantly lower pain intensity than the group receiving physiotherapy with on-line education. ⁴¹ The Kaia App was evaluated previously in a retrospective cohort study and showed good effect on pain intensity reduction, ⁴³ and the analysis of the updated version of Kaia App revealed improvement in treatment adherence. ⁴² One more pilot RCT confirmed the feasibility and preliminary effectiveness of a physical activity intervention supported by a smartphone app on pain and disability. At the same time, patient acceptance and reduced care-seeking were observed. ³⁶ The protocol of this study is detailed in a separate article. ⁴⁵ Other studies focused on qualitative analysis of the patient’s preferences and compliance in the smartphone app use ^{35 , 39} and case series analysis of feasibility, effectiveness on pain and exercices adherence, and patient satisfaction with the telerehabilitation booster session. ³⁷ Selter et al., in their pilot study, assessed the validity of an image-based quantification of pain-related disability as well as patient compliance and patient-perceived utility of the smartphone app ( Limbr). ⁴⁴ The results showed good compliance and patient-perceived utility. In addition, the authors noted good validity of an image-based quantification of pain-related disability. ⁴⁴

The most-studied use of the smartphone app in LBP patients was for self-management. We identified 29 articles focused on self-management of LBP via a smartphone app: four systematic reviews, ^{10 , 33 , 46 , 47} five RCTs, ^{11 , 38 , 41 , 48 , 49} five study protocols, ^{18 , 45 , 50 – 52} six qualitative studies, ^{35 , 39 , 53 – 56} three retrospective cohort studies, ^{26 , 42 , 43} five pilot RCTs, ^{36 , 44 , 57 , 58} and one case series. ³⁷ Many of the studies coupled the self-management and telerehabilitation in the same app or in the same study intervention. These studies were presented in the previous section. ^{32 , 35 – 45} In a scoping review of the outcome used in the smartphone apps for self-management of LBP, Coe-O’Brien et al., found 74 apps; only four used the outcome measure that could be linked to the International Classification of Functioning, Disability and Health System (ICF) core set for LBP. Furthermore, they concluded that most of the apps were of low quality, indicating the lack of outcome evaluation in the apps. ³³ Machado et al., performed a systematic review of smartphone apps for self-management of LBP. Similarly, they analysed 61 apps and found overall low quality of the apps, pointing to the lack of studies evaluating their effectiveness, presentation of the questionable information, and unattractive layouts. They also recommend that app developers collaborate with healthcare professionals and researchers to ensure the benefit for LBP patients. ¹⁰ A recent systematic review with meta-analysis analysed eHealth based self-management for chronic LBP. The term eHealth encompassed the interventions based on smartphone apps (m-Health) or on traditional Internet (web-Health) use, so studies included in this meta-analysis were not all based on smartphone apps. Nevertheless, the authors concluded low to moderate evidence of a positive impact on pain and disability of this type of self-management. ⁴⁶ Previously Nicholl et al., performed a systematic review of digital support interventions for the self-management of LBP. They revealed heterogeneity and low quality of the studies, which could not support the utility of digital based interventions for LBP. ⁴⁷

A summary of the RCTs focused on effectiveness of smartphone apps targeting self-management for LBP is presented in Table 2 as is the quality evaluation using the PEDRO score. Chhabra et al., evaluated the effect of the self-management app Snapcare. A control group of LBP patients received a written prescription of medication and home exercises and were compared to a self-management group focusing on physical activity improvement based on Snapcare. The results at 12 weeks showed a similar improvement in pain in both groups and greater improvement in function in the Snapcare than control group. The authors concluded that such results support the utility of Snapcare for LBP patients. ¹¹ Suman et al., performed a cluster RCT evaluating the effectiveness and cost-utility of a multifaceted eHealth intervention based on websites and social media platforms but also including a mobile version, adaptable to a smartphone. The intervention was inspired by the Australian mass media campaign promoting physical activity, positive back beliefs and coping with LBP. The control group received a digital patient information letter. The presented eHealth strategy was not effective in improving patients’ back pain beliefs or decreasing disability or absenteeism, but the study provided a favourable cost-utility analysis. ⁴⁸ Irvine et al., showed that their mobile app FitBack was effective in improving physical, behavioural and worksite outcomes. FitBack was based on a self-tailored cognitive-behavioural approach and used the American Pain Society recommendations. ⁴⁹ Almhdawi et al., also reported that use of their app called Relieve my back is efficient in pain and disability self-management. ⁵⁸

We also found several articles presenting study protocols for smartphone apps used by LBP patients. Sandal et al., presented the protocol of an RCT comparing the effect of usual care supported by the selfBACK app versus usual care only. Tailored self-management plans were provided by the selfBACK app consisting of advice on physical activity, physical exercises, and educational content. Self-management plans were prepared by using case-based reasoning methodology, a branch of artificial intelligence. ⁵⁰ Mork et al., presented complementary information regarding the implementation of the methodology used in the selfBACK protocol. ¹⁸ One RCT protocol aimed to evaluate the effectiveness of Relaxback, focusing on relaxation for LBP patients in reducing pain. Autogenic training, mindfulness meditation and guided imagery is used in the app and will be compared to usual care. ⁵¹ Another RCT protocol is for MIMate, designed to support motivational interviewing performed by a physical therapist and targeting behavioural changes regarding physical activity. MIMate is used between face-to-face physical therapy sessions and is compared to usual physical therapy sessions. ⁵²

Two qualitative studies coupled telerehabilitation and self-management interventions and were mentioned in the previous section. ^{32 , 36} Nevertheless, four other qualitative studies investigated the preference in content of an app improving self-management ^{53 , 55} or the feasibility/utility of an app providing educational videos focused on self-management, postoperative protocols, or tailored self-management plans. ^{54 , 56} Three retrospective cohort studies using smartphone apps for self-management of LBP were coupled with telerehabilitation, evaluation or data collection and were mentioned in the previous sections. ^{26 , 42 , 43} Similarly, in two pilot studies, self-management was coupled with telerehabilitation; these articles were presented in the previous section. ^{36 , 44} However, another pilot RCT, used a smartphone app to enhance self-management between physical therapy sessions and compared it to physical therapy only. The Pain Care app provided self-management based on home exercises. The authors concluded that a more powerful study needed to be conducted considering their promising results. ⁵⁷

Some studies used smartphone apps for LBP patients to collect medical information. One study protocol presented a smartphone app to collect timely data in ecological situations, including pain intensity, physical function, analgesic use and adherence to auricular point acupressure treatment. ⁵⁹ Similarly, in a study that aimed to evaluate relations between opioid use and pain intensity, a smartphone app was used exclusively to collect information about pain intensity multiple times daily. ⁶⁰ However, in most studies, data were collected together with other uses of the smartphone app previously presented: self-management, telerehabilitation, evaluation. ^{23 – 26 , 42 , 43 , 50 , 54}

Some articles could have been missed in our search strategy as we only used three databases. However, based on a large problem of our review, identifying the current fields of the smartphone apps use for LBP patients’ points out perspectives for further studies including perspectives for a more specific systematic review. Also, the highly heterogenic terminology regarding smartphone apps use could have resulted in some omissions. Several studies using eHealth strategy ( e.g., tablet apps) were not included if there was no clear information about the smartphone app use.

All data underlying the results are available as part of the article and no additional source data are required.