This review will include randomized and nonrandomized studies. For nonrandomized studies, case–control, cohort, and cross-sectional studies in which MHCS was the main primary intervention strategy used for chronic disease awareness will be included. Studies in which MHCS was used as an intervention strategy for chronic disease screening will also be included.

Study participants will be both female and male, and no age restriction will be applied. Individuals of any race, ethnicity, employment status, occupation, and geographical location will be eligible for inclusion.

The relevant MHCS interventions for this study focus primarily on positively impacting chronic diseases awareness. MHCS interventions for health care consumers vary. For example, in the study by Pryss et al [16], MHCS was designed to provide patients with aggregated information about the variation of their tinnitus over time. This is a form of health awareness strategy where the mobile feedback service helps a patient to understand, get better control of, and cope with their chronic health condition. Meanwhile, he Free et al’s [20] strategy categorizes by device (eg, mobile phone, personal digital assistant [PDA]) and modality (eg, SMS, text messaging, multimedia message service [MMS], video). This will be used in describing the interventions [20].

Mobile devices having sensory capabilities, such as GPS, camera, and light, is a typical feature of the technology used for MHCS. Also, cellular communication that allows for wireless and 3G/4G capabilities will be part of the study. These devices include mobile phones (including Android and IOS smartphones), PDAs and PDA phones, tablets, smartphone apps, ultraportable computers, and smart books [10,21].

MHCS smartphone functionalities comprise voice over internet protocol, SMS, text messaging, MMS, GPS, Bluetooth, audio, email, light sensing, and internet [22]. Some applications using MHCS modalities include data collection (web) [23,24], patient health education (web) [6], patient health self-management (text) [25,26], behaviour change communication (images, audio, text) [27], sensors and point-of-care devices (camera, microphone, GPS, accelerometer, digital compass, Bluetooth sensing) [28], provider communication (SMS, MMS, smartphone camera), provider training and education (SMS, MMS, audio, video), human resource management (voice, SMS), supply chain management (GPS, SMS), and financial transactions and incentives (mobile banking service, airtime transfers) [10,17,29].

The proposed MHCS intervention impact on chronic disease awareness will be reviewed by assessing the following: the increased attendance at clinics for chronic diseases; the stage of chronic diseases when diagnosed (as this will assist in determining whether MHCS has promoted early detection and screening); and increased chronic disease inquiry via call centres, online forums, and social media. User acceptability will not be assessed as an outcome.

Chronic disease awareness is described as the ability to be fully informed and knowledgeable of a terminal disease suffered by a patient [30]. Furthermore, where actual patient numbers cannot be determined, the baseline for assessment will be determined by the keywords increase, improvement, or, rise used in the study.

Geographical setting will not be restrictive in the study. All available health facilities where MHCS research on chronic diseases were conducted will be included. This approach allows for all relevant information sources to be captured.

Study types will be excluded according to the following criteria described by Tokosi et al [25]: non–English language papers; studies before January 1, 1999, and after December 31, 2019; nonhealth-related studies; letters, commentaries, and editorials; studies lacking primary data or explicit method description; duplicate studies that are published in more than one journal or report (the most comprehensive and up-to-date version will be used); studies not including human involvement; and studies not having a health-technology application focus.

Banda [2] highlights the origins of crowdsourcing that date back to the 19th century when Joseph Henry created the first national weather map of the United States in 1856 using a new networked technology of his day (the telegraph) to crowdsource weather reports from across the country [31]. In our study, the earliest start date will be when each database began operation. All major databases, including MEDLINE, Embase, PsycINFO, CINAHL, The Cochrane Library (Cochrane Database of Systematic Reviews, Cochrane Central Register of Controlled Trials, Cochrane Methodology Register), National Health Services Health Technology Assessment Database, Scopus, Web of Science, and Google Scholar, will be included [30]. The language of publication will be limited to English for reasonable analysis purposes.

The phrase chronic disease will be used to expand the search strategy (where necessary), which will include grey literature and other databases, including SpringerLink, Wiley InterScience, trial registers, Institute of Electrical and Electronics Engineers, Association for Computing Machinery Digital Library, and CiteSeer [29]. We concur with the experimental findings proposed by Fortuin et al [10] in identifying accurate search terms in the development of an optimum search strategy. For trial registers, we will identify ongoing studies and recently completed trials. The studies to be included will be selected using predefined search terms adapted for the databases to be used. Full-text articles of studies extracted from reference lists will be reviewed. Manual searches of reference lists of primary studies, and of relevant and previously published reviews will be completed. Grey literature will comprise unpublished studies identified in universities and other institutional repositories, with the same eligibility criteria applied. The MEDLINE format for searching key terms is detailed in Table 1 and in Multimedia Appendix 2. The number of identified references following a preliminary search is reported in Table 1.

TT will retrieve all relevant articles from various databases, based on the finalized search strategy. All the literature obtained will be saved in Endnote reference management software (Clarivate Analytics). Both authors will independently screen the titles and abstracts of retrieved studies for eligibility. Both authors will make a final assessment for inclusion using the full-text article. Discrepancies and disagreements will be resolved by both authors.

A standardized data extraction form adapted from a study by Tokosi et al [30] will be used for data extraction. Full texts of selected abstracts will be retrieved and data extracted in line with the prespecified template. The key information to be extracted includes the following: author name(s) and year of the study, type of participant/study population and demographic characteristics; type of mHealth device used (eg, mobile phones, PDAs, smartphones, tablets); type of intervention (eg, SMS, MMS, video, text, audio); nature of the mHealth intervention (eg, awareness, diagnosis, treatment); type of study (eg, randomized, nonrandomized); type of outcome measured; and findings/results.

Following Saidi et al’s [32] procedure, data will be entered into Review Manager software, version 5.3. (Cochrane). Both authors will verify the data entered for missing or incorrect data. If both authors cannot agree on missing data entry, an independent third party will be consulted to mediate. Missing data will be requested from study authors via email [21]. If no response is received from study authors, an attempt will be made to impute missing SD or standard error values using data from other similar studies in the review with similar methods and sample sizes, as recommended by Wiebe et al [33].

The authors will use recommendations by the International Cochrane Collaboration to independently assess the risk of bias [34]. These criteria include randomization sequence generation, treatment allocation concealment, blinding of participants, incomplete outcome data, selective outcome reporting, and other sources of bias [30]. All included studies will be scored for bias using these criteria. A descriptive summary for each scoring will be recorded. Discrepancies between the review authors regarding the risk of bias in particular studies will be resolved by dialogue, with involvement of an independent third party, where necessary.

The extracted data will be presented in an evidence table adapted from Saidi et al’s [32] study. A descriptive synthesis will be undertaken in accordance with the Centre for Reviews and Dissemination [35]. Continuous outcomes will be ascertained by calculating mean differences and SDs. Ratios and their corresponding 95% CIs will be determined for dichotomous outcomes. Heterogeneity will be used to examine participants, interventions, and outcomes of each study. The statistical test for heterogeneity will include the I² test which quantifies heterogeneity; this test will allow for the quality of evidence to be validated [34]. Data will be pooled; where collected data are sufficiently similar, a meta-analysis will be conducted. Similarly, where the variability between studies is high, the results will not be pooled and a narrative synthesis will take place [30]. When appropriate, a subgroup analysis will be used to determine if varying mHealth crowdsensing applications have an impact on chronic disease awareness and in what context this occurs. Subgroups to be considered for this analysis will include age grouping and geographical region.

Various sensitivity analyses as espoused by Tokosi et al [30] will be performed, including analysis conducted based on the study quality (risk of bias and level of participant dropout) to investigate possible sources of heterogeneity. Another analysis will be used to determine how excluded studies could have influenced the overall result. A final analysis to determine how the result would differ from other study results should there be only high-quality studies included will be used [34].