Skip to main content
NCBI home page
Scandinavian Journal of Primary Health Care logoLink to Scandinavian Journal of Primary Health Care
. 2026 Apr 25;44(1):2656694. doi: 10.1080/02813432.2026.2656694

Overdiagnosis in atrial fibrillation screening with wearables

Christoffer Bjerre Haase a,b,✉, Alma Emilie Jensen a, Frederikke Agerbo Modin a, Volkert Siersma a, John Brandt Brodersen c
PMCID: PMC13112866  PMID: 42033454

Abstract

Purpose

Wearable technologies are increasingly used to screen for atrial fibrillation (AF), often outside formal clinical pathways. While this holds promise for early detection, the impact on patient outcomes remains unclear. We aimed to examine the hypothesis whether intermittent and continuous AF screening using wearable-like technologies leads to overdiagnosis in asymptomatic, high-risk individuals.

Methods

This hypothesis-generating analysis draws on available RCTs screening asymptomatic, high-risk individuals for AF using wearable-like ECG technologies. It is not a systematic review, as the field remains too early-stage for formal evidence synthesis. Eligible studies included either intermittent (≥10 recordings) or continuous monitoring for ≥1 week, reflecting typical smartwatch use. Primary outcomes included AF diagnoses, anticoagulant use, and adverse events. Overdiagnosis was inferred when excess AF detection did not translate into improved clinical outcomes. Analyses used Mantel-Haenszel random-effects models.

Results

Continuous screening nearly tripled AF diagnoses (OR 2.84; 95% CI, 1.61–5.02) and doubled anticoagulant use compared to controls. Intermittent screening showed a non-significant 1.8-fold increase. Neither method significantly reduced adverse outcomes. A pooled analysis revealed a marginal reduction in negative events (OR 0.94, 95% CI 0.90–0.99). Overdiagnosis estimates varied by method, ranging from 8–29% based on persistent differences in cumulative incidence after the screening period has ended and 47–65% based on increased diagnoses and treatment during the screening phase that are not followed by improved patient-relevant outcomes.

Conclusion

Wearable-enabled screening for AF in high-risk patients increases diagnosis and treatment without clear benefit in clinical outcomes. These findings support the hypothesis of overdiagnosis and findings highlight the need for cautious integration of consumer technologies into primary care. Further trials with extended follow-up in low-risk populations are needed to clarify when AF detection improves outcomes versus when it leads to unnecessary medicalisation and harm.

Keywords: Atrial fibrillation, wearables, overdiagnosis, screening, low-value care

Introduction

Wearable devices are increasingly viewed as promising tools in primary care for both in-person and virtual care delivery [1]. These advancements have enabled the screening of cardiac arrhythmias, particularly atrial fibrillation (AF), with the aim of preventing stroke. Yet, growing access to private and home-based screening also shifts the burden of overdiagnosis, false positives and follow-up care onto general practice, raising concerns about cost, responsibility, and clinical value [2]. AF is the most prevalent cardiac arrhythmia and, when diagnosed clinically, is associated with a fivefold increase in stroke risk, often leading to severe complications and long-term disability [3]. The estimated global prevalence of AF is 2–4%, a figure projected to double by 2050 due to an aging population and increasing prevalence of risk factors such as diabetes and hypertension [4].

Given that 20–30% of all strokes are associated with AF, screening in general practice for AF has been proposed as a preventive measure [5]. However, the 2020 guidelines from the European Society of Cardiology (ESC) [6] and the U.S. Preventive Services Task Force (USPSTF) [7] do not recommend systematic AF screening for any population due to insufficient evidence of the benefits outweighing the harms. Despite these recommendations, an increasing number of individuals are engaging in self-screening outside traditional healthcare systems by using wearable technologies such as smartwatches [8]. General practitioners (GPs) are reporting how their patients bring these data to the consultation with the ex­­pectations of diagnostic assessments and potential clinical consequences [9]. While wearables and AI create opportunities to redefine primary care delivery in more patient-centred, location-flexible ways [10], the optimism among GPs is also followed by concerns about negative changes in workflows and prioritizations [11].

The critical question is whether these technologies improve diagnostic accuracy and patient outcomes or, conversely, lead to overdiagnosis that increases the burden on individuals and healthcare systems by detecting subclinical arrhythmias [12]. We hypothesise that continuous screening for AF using wearable technologies substantially increases AF prevalence without leading to meaningful improvements in patient-relevant outcomes, thereby contributing to overdiagnosis.

This paper is not a systematic review but a hypothesis-generating analysis based on the best available trial data. We deliberately avoid overstating the strength or generalisability of the findings, as the research field of wearable-enabled AF screening is still too early in development to support formal evidence synthesis.

Investigating this question thoroughly requires large-scale, high-quality RCTs with extended follow-up in low-risk populations. However, such studies remain limited, despite growing commercial interest. Most wearables are marketed as wellness products [13,14] and are therefore not subject to the regulatory demands applied to medical devices, screening programmes, or diagnostic tools [15]. As a result, rigorous research on their long-term clinical implications is lacking.

Closing this knowledge gap is therefore essential for informing future research and policymaking, particularly in general practice, where GPs are often the first to assess the consequences of patient-generated wearable data. Based on the best available evidence of AF screening, we therefore examined whether intermittent and continuous AF screening increases detection rates and whether those diagnosed through these methods experience improved or unchanged prognoses

Diagnosis and screening recommendation of atrial fibrillation

Diagnosing AF poses significant challenges, as the prevalence of asymptomatic AF is estimated at approximately 27%, though this varies widely depending on the distribution of associated risk factors [16]. The current clinical standard for AF diagnosis relies on a 12-lead ECG or a single-lead ECG recording lasting more than 30 s. This method frequently fails to detect intermittent arrhythmic episodes, known as paroxysmal AF, due to the limited duration of recording [5].

Although the ESC and USPSTF do not endorse systematic AF screening, the ESC recommends opportunistic screening in individuals aged ≥65 years using pulse palpation or an ECG rhythm strip [6]. Additionally, the ESC suggests that ‘systematic ECG screening should be considered to detect AF in individuals aged ≥75 years or those at high risk of stroke’ (p.395) [6]. In contrast, the USPSTF acknowledges the potential benefits of AF screening but does not recommend widespread implementation because of insufficient evidence that the benefits outweigh potential harms [7]. Despite these guidelines, various stakeholders—including healthcare policymakers, national health services, scholars and patient advocacy groups—continue to advocate for increased AF screening [17].

Overdiagnosis

Overdiagnosis is defined as the identification of conditions that would never cause harm, or as the medicalisation of normal life experiences due to expanded disease definitions [18]. The increasing use of big data in screening programmes may exacerbate this issue, leading to side effects of treatment and psychosocial stigma from being diagnosed, as well as waste of resources from unnecessary tests and treatments [19].

Screening for AF with wearable technologies

Wearable devices, also referred to as mobile health (mHealth) technologies, are frequently described prospectively as heralding a ‘medical revolution’ [20]. These devices offer continuous monitoring of physiological parameters, allowing real-time data collection that can be shared with healthcare providers. Still, in primary care, research is lacking to evaluate the real-world clinical impact of wearables and other AI applications [21]. In a primary care population, Portable ECG Monitors [22], smartphone-operated single-lead ECG devices [23], and AI algorithms applied to electronic health records [24] have shown promising accuracy in detecting atrial fibrillation. However, the extent of overdiagnosis has not been investigated.

Regarding actual AF screening rather than AF detection, few studies have investigated the effectiveness of wearables despite their increasing popularity. None have systematically assessed the extent of overdiagnosis. Current policy discussions on how to use wearable data for diagnostic purposes in general practice have been criticised as overly optimistic and insufficiently evidence based [9].

Methods: How can overdiagnosis from atrial fibrillation screening Be measured?

A search was conducted in PubMed to identify RCTs evaluating AF screening in asymptomatic individuals using technologies similar to wearable ECG patches and at-home pulse monitoring devices. Studies were included if they assessed populations without a prior AF diagnosis and if participants had an elevated risk of stroke or AF. Additional inclusion criteria required studies to have at least 10 ECG recordings (categorised as intermittent) or continuous monitoring (categorised as continuous) for a minimum of one week. We required ≥10 ECG recordings for intermittent screening and ≥1 week of continuous monitoring to reflect realistic use patterns of smartwatches and similar consumer devices. Manually triggered recordings are unlikely to be repeated unless symptoms prompt concern, making ≥10 recordings a reasonable proxy for engaged use. Likewise, wearables with passive rhythm detection are typically worn for extended periods, justifying a one-week minimum. These thresholds ensure that included studies mirror real-world wearable use while maintaining methodological relevance. Given the limited availability of high-quality RCTs on wearables, we included studies that examined various screening methods and age groups, to provide a broader perspective on the topic.

Exclusion criteria were prior AF diagnosis, previous initiation of anticoagulation therapy, prior stroke or embolism, and studies assessing the efficacy of specific screening tools rather than the impact of screening itself. Primary outcomes were all-cause mortality, stroke/transient ischemic attack (TIA), systemic embolism, bleeding events, combination of the mentioned events (see details in Table 1), AF diagnoses during and after the study period, and anticoagulant use. Meta-analyses were conducted using the Mantel-Haenszel method to obtain a random-effect meta-analysis odds ratio (OR), and results were visualised in forest plots [25].

Table 1.

Meta-analyse or with 95% CI.

  Continuous screening Intermittent screening All studies
AF diagnoses during study period 2·84 (1·61; 5·02)
I2 = 94%		 1·88 (0·55; 6·40)
I2 = 84%		 2·48 (1·46; 4·23)
I2 = 98%
AF diagnoses after study period 1·40 (1·19; 1·65)
I2 = 40%		 1·09 (1·02; 1·17)
I2 NA		 1·27 (1·05; 1·55)
I2 = 84%
Anticoagulant treatment 2.·06 (1·39; 3·07)
I2 = 89%		 1·84 (0·51; 6·62)
I2 =85%		 1·93 (1·27; 2·93)
I2 =97%
All-cause death 0·94 (0·82; 1·09)
I2 = 0%		 0·96 (0·91; 1·01)
I2 = 0%		 0·96 (0·91; 1·01)
I2 = 0%
Stroke/TSI 0·93 (0·66; 1·33)
I2 = 32%		 0·91 (0·83; 1·00)
I2 = 0%		 0·91 (0·83; 1·00)
I2 = 0%
Systemic emboli 0·75 (0·32; 1·76)
I2 NA		 0·94 (0·34; 2·57)
I2 = 17%		 1·03 (0·73; 1·43)
I2 = 0%
Bleeding 1·07 (0·74; 1·53)
I2 = 57%		 0·99 (0·92; 1·07)
I2 = 0%		 1·02 (0·89; 1·16)
I2 = 9%
Combined 0·91 (0·81; 1·03)
I2 =0%		 0·95 (0·90; 1·00)
I2 = 0%		 0·94 (0·90; 0·99)
I2 = 0%
Open in a new tab

Various odds ratio for continuous and intermittent screening as well as total.

Combined means the pooled numbers of all the events measured within the specific study.

Svennberg 2021: stroke, systemic embolism, major bleeding, and death.

Halcox 2017: stroke, TIA, systemic embolism, bleeding, and death.

Steinhubl 2018: stroke, MI, emboli, and death.

Gladstone 2021: death, emboli, stroke, TSI, and death.

Svendsen 2021: stroke, systemic arterial embolism, major bleeding, and death.

Lopes 2024: stroke, bleeding, and death.

Overdiagnosis of AF is conventionally measured from a persistently higher cumulative AF incidence in the screened group compared to the control group after the end of the screening period and following adequate long-term follow-up. This indicates that the additional cases detected through screening do not represent earlier detection of clinically relevant AF that would have otherwise emerged, but rather detection of cases unlikely to cause harm (i.e. lead-time without benefit) [14,26–29]. Another way overdiagnosis presents itself is through increased anticoagulant treatment without a corresponding reduction in adverse clinical outcomes (e.g. stroke, systemic embolism, mortality, or major bleeding). This suggests that the newly diagnosed AF cases led to treatment but did not improve patient-relevant outcomes which are expected to happen quickly after the appearance of (maybe undetected) AF. Overdiagnosis is then measured from a higher cumulative AF incidence in the screened group compared to the control group after the end of the screening period alone [30].

Hence, we applied two complementary methods to estimate overdiagnosis from the meta-analysis ORs:

  • The Post-Screening Divergence Method, where overdiagnosis is calculated as (1–1/OR) x 100%, using the odds ratios for AF diagnoses during and after the screening period. Here we can only use the data from the studies that report follow-up beyond the screening period.

  • The Treatment–Outcome Discrepancy Method, where overdiagnosis is also calculated as (1–1/OR) x 100%, but now using the odds ratios for AF diagnoses only during the screening period; while at the same time observing that also anticoagulant treatment increases in a similar magnitude, but there is no decrease in adverse clinical outcomes.

Results: continuous screening led to increased diagnosis of atrial fibrillation with uncertain beneficial effects

Six RCTs evaluating AF screening in asymptomatic, high-risk individuals using wearable-like ECG technologies were identified. Three studies included follow-up periods extending beyond the screening phase, allowing for an assessment of overdiagnosis.

Continuous screening was associated with a nearly threefold increase in AF diagnosis rates compared to control groups (Figure 1). Additionally, anticoagulant initiation was twice as likely in the screened group (Table 1). In contrast, intermittent screening demonstrated a non-significant trend toward increased AF detection and anticoagulant use, with an approximately 1·8-fold increase compared to controls.

Figure 1.

Forest plot displaying odds ratios and confidence intervals for studies on intermittent and continuous technology. A forest plot visualizes odds ratios (OR) and 95% confidence intervals (CI) for studies comparing intermittent and continuous technologies. The left section shows studies by authors, ORs, CIs, and weights for intermittent (Svennberg 2021, Halcox 2017) and continuous technologies (Steinhubl 2018, Gladstone 2021, Svendsen 2021, Lopes 2024). Total ORs for intermittent and continuous are 1.88 and 2.84. Heterogeneity statistics are provided. The right panel graphs these ORs with squares and horizontal lines for CIs, including a vertical line at OR=1 indicating no effect. Overall pooled results yield an OR of 2.48.

Open in a new tab

Meta-analysis of AF-diagnoses based on screening technique.

The chance of being diagnosed with AF in intermittent or continuous screening, and total, compared to the control groups

Neither continuous nor intermittent screening demonstrated significant reductions in adverse clinical outcomes. However, a pooled analysis of both screening strategies showed a marginally significant reduction in negative outcomes (OR: 0·94, 95% CI: 0·90–0·99).

Using the Post-Screening Divergence Method, overdiagnosis was estimated at 8% for intermittent screening and 29% for continuous screening. Using the Treatment–Outcome Discrepancy Method, overdiagnosis was estimated at 47% for intermittent screening and 65% for continuous screening.

Discussion

Our analysis found that continuous screening for atrial fibrillation (AF) in high-risk populations nearly tripled diagnosis rates and doubled anticoagulant initiation compared to controls. However, no significant reduction in adverse outcomes was observed. Intermittent screening showed a smaller, non-significant increase in diagnoses and treatment. These findings support the hypothesis that wearable-based AF screening may lead to overdiagnosis, with limited clinical benefit.

Limitations of the method

Our proposed method supports the hypothesis of potentially significant overdiagnosis of AF by using wearable-like technologies. However, further RCTs with extended follow-up periods in general populations are required to fully assess the potential benefits and harms of AF screening with wearables in low-risk individuals. Only three studies included follow-up periods beyond the screening phase, with durations of 2·5 years [31], 3 years [32], and 7 years [33]. Additionally, four of the six studies lacked participant blinding [3,31,34,35]. The included studies primarily focused on older populations with an elevated AF risk, limiting generalisability to younger, lower-risk individuals. Heterogeneity in study methodologies, reflected in high I2 values for several results, further complicates interpretation of the data.

These limitations significantly affect how the results should be interpreted. The fact that only three studies extended follow-up beyond the screening period limits our ability to assess whether early AF detection improves long-term clinical outcomes or simply reflects lead-time bias. Without sustained follow-up, it is unclear whether additional diagnoses translate into benefit.

The lack of participant blinding in most studies introduces risks of performance and detection bias, especially regarding treatment decisions such as anticoagulation, which may be influenced by knowledge of screening assignment.

Moreover, the focus on older, high-risk populations limits external validity. As wearables are widely used by younger, lower-risk individuals, the benefits and harms observed in these trials may not generalise. This raises the possibility of spectrum bias, where the diagnostic and clinical consequences differ by baseline risk.

Finally, substantial methodological heterogeneity, reflected in high I2 values, limits the reliability of pooled estimates and weakens confidence in overarching conclusions. Together, these issues highlight the need for well-designed, long-term trials in more diverse populations to clarify the real-world implications of wearable AF screening.

Implications: What to do from there

Our analysis of existing data revealed a high degree of overdiagnosis as well as uncertainty of any benefits in high-risk individuals when they are continuously screened for AF for a longer period. Given the hypothesis we examine, applying these findings to low-risk populations—such as younger wearable users in general practice—would likely yield an even less favourable benefit-to-harm ratio. However, a systematic review and meta-analysis of the benefits and harms of AF screening based on RCTs with longer follow-up periods is needed.

While wearable screenings for AF may provide benefits for some individuals, their overall utility from a public health standpoint remains questionable. In the Apple Heart Study, 2,161 participants out of 419,297 participants (0.52%) received notifications of irregular pulse [36]. Of these 2,161 participants, 450 were successfully monitored and analysed with additional ECG patches. Ultimately, 34% of those monitored (153 individuals) were diagnosed with AF. This indicates a significant involvement from citizens and healthcare systems to diagnose relatively few individuals, without certainty of benefit and in line with the hypothesis of overdiagnosis.

Current political priorities have been criticised as overly optimistic and lacking evidence-based support9. As mentioned, the wearables are often categorized as wellness technologies with sparse regulation, and an international comparison of health app policies among different countries has shown little effort to regulate the health app market [37]. The world’s largest technology and data companies are producing wearable devices, viewing healthcare as a crucial new market for expansion [38]. Policymakers could implement regulations that incentivise manufacturers to fund independent testing, enabling their products to gain accreditation as valid diagnostic tools within healthcare systems.

For now, our findings suggest that the rise of wearable health devices is likely contributing to increased AF overdiagnosis. Since wearable users may often be relatively young [39,40], it is important in general practice to consider whether these users fall within the main group of concern of AF, as defined by the ESC and USPSTS.

Conclusion

Our findings support the hypothesis that continuous AF screening in high-risk populations nearly triples AF diagnosis rates without significantly improving patient outcomes, suggesting substantial overdiagnosis. However, this remains a hypothesis in the general population, especially low-risk individuals, and requires further investigation. The long-term clinical impact of AF detected through continuous wearable monitoring in low-risk individuals is still unknown, and the balance between benefits and harms has not been adequately explored in current research.

This underscores the urgent need for dedicated RCTs with extended follow-up periods in low-risk populations to confirm or refute this hypothesis. Future studies should focus on determining when AF requires medical intervention and when it may represent a benign condition that does not necessitate treatment. As wearable technologies become increasingly integrated into healthcare, it is essential that evidence-based research guides their implementation, rather than commercial or technological momentum alone. A more nuanced understanding of the implications of wearable-driven AF screening is needed to ensure that screening strategies do not contribute to unnecessary medicalisation, overtreatment, or an increased burden on healthcare systems.

Acknowledgements

None.

Disclosure statement

JBB is a member of the board of the Overdiagnosis Conference with international partners such as the BMJ plus Dartmouth, Bond and Oxford universities based on a non-profit model. The remaining authors declare no conflicts of interest.

Support

No support for this work

Prior presentation(s)

None

References

  • 1.Sawler S, Reid R, Kolen A.. Understanding primary health care provider’s perceptions of using activity monitors: a qualitative study. The Annals of Family Medicine. 2023;21(Supplement 1):4224. doi: 10.1370/afm.21.s1.4224. [DOI] [Google Scholar]
  • 2.Robson R. Who should pay for reviewing the ECGs from the Apple Watch 4 series? Br J Gen Pract. 2019;69(684):333–333. doi: 10.3399/bjgp19X704249. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Svendsen JH, Diederichsen SZ, Højberg S, et al. Implantable loop recorder detection of atrial fibrillation to prevent stroke (The LOOP Study): a randomised controlled trial. Lancet. 2021;398(10310):1507–1516. doi: 10.1016/S0140-6736(21)01698-6. [DOI] [PubMed] [Google Scholar]
  • 4.Krijthe BP, Kunst A, Benjamin EJ, et al. Projections on the number of individuals with atrial fibrillation in the European Union, from 2000 to 2060. Eur Heart J. 2013;34(35):2746–2751. volume14 September Pages doi: 10.1093/eurheartj/eht280. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Robson J, Schilling R.. Atrial fibrillation: time for active case finding. Br J Gen Pract. 2019;69(679):58–59. doi: 10.3399/bjgp19X700985. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Hindricks G, Potpara T, Dagres N, et al. 2020 ESC Guidelines for the diagnosis and management of atrial fibrillation developed in collaboration with the European Association for Cardio-Thoracic Surgery (EACTS): the Task Force for the diagnosis and management of atrial fibrillation of the European Society of Cardiology (ESC) Developed with the special contribution of the European Heart Rhythm Association (EHRA) of the ESC. Eur Heart J. 2021;42(5):373–498. doi: 10.1093/eurheartj/ehaa612. [DOI] [PubMed] [Google Scholar]
  • 7.Davidson KW, Barry MJ, Mangione CM, et al. Screening for atrial fibrillation: US preventive services task force recommendation statement. JAMA. 2022;327(4):360–367. doi: 10.1001/jama.2021.23732. [DOI] [PubMed] [Google Scholar]
  • 8.Ponamgi SP, Siontis KC, Rushlow DR, et al. Screening and management of atrial fibrillation in primary care. BMJ. 2021;373:n379. doi: 10.1136/bmj.n379. [DOI] [PubMed] [Google Scholar]
  • 9.Haase CB, Ajjawi R, Bearman M, et al. Data as symptom: doctors’ responses to patient-provided data in general practice. Soc Stud Sci. 2023;53(4):522–544. doi: 10.1177/03063127231164345. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Hansmann KJ, Chang T.. Defining the “New Normal” in Primary Care. Ann Fam Med. 2021;19(5):457–459. doi: 10.1370/afm.2711. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Moschogianis S, Darley S, Coulson T, et al. Seven opportunities for artificial intelligence in primary care electronic visits: qualitative study of staff and patient views. Ann Fam Med. 2025;23(3):214–222. doi: 10.1370/afm.240292. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Fitzpatrick NP. Asymptomatic atrial fibrillation patients: DOAC or don’t? Br J Gen Pract. 2025;75(755):254–254. doi: 10.3399/bjgp25X742497. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Sharon T. Blind-sided by privacy? Digital contact tracing, the Apple/Google API and big tech’s newfound role as global health policy makers. Ethics Inf Technol. 2021;23(Suppl 1):45–57. doi: 10.1007/s10676-020-09547-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Sharon T. From hostile worlds to multiple spheres: towards a normative pragmatics of justice for the Googlization of health. Med Health Care Philos. 2021;24(3):315–327. doi: 10.1007/s11019-021-10006-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Kalarus Z, Mairesse GH, Sokal A, et al. Searching for atrial fibrillation: looking harder, looking longer, and in increasingly sophisticated ways. An EHRA position paper. Europace. 2023;25(1):185–198. doi: 10.1093/europace/euac144. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Pamporis K, Karakasis P, Sagris M, et al. Prevalence of asymptomatic atrial fibrillation and risk factors associated with asymptomatic status: a systematic review and meta-analysis. Eur J Prev Cardiol. 2025:1–12. doi: 10.1093/eurjpc/zwaf138. [DOI] [PubMed] [Google Scholar]
  • 17.Mahase E. Atrial fibrillation: most UK media “promote screening” despite lack of evidence, study finds. BMJ. 2022;379:o2449. doi: 10.1136/bmj.o2449. [DOI] [PubMed] [Google Scholar]
  • 18.Brodersen J, Schwartz LM, Heneghan C, et al. Overdiagnosis: what it is and what it isn’t. BMJ Evid Based Med. 2018;23(1):1–3. doi: 10.1136/ebmed-2017-110886. [DOI] [PubMed] [Google Scholar]
  • 19.Vogt H, Green S, Ekstrøm CT, et al. How precision medicine and screening with big data could increase overdiagnosis. BMJ. 2019;366:l5270. doi: 10.1136/bmj.l5270. [DOI] [PubMed] [Google Scholar]
  • 20.Young AJ. New technologies and general practice. Br J Gen Pract. 2016;66(653):601–602. doi: 10.3399/bjgp16X688021. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Kueper JK, Terry AL, Zwarenstein M, et al. Artificial intelligence and primary care research: a scoping review. Ann Fam Med. 2020;18(3):250–258. doi: 10.1370/afm.2518. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Kristensen AN, Jeyam B, Riahi S, et al. The use of a portable three-lead ECG monitor to detect atrial fibrillation in general practice. Scand J Prim Health Care. 2016;34(3):304–308. doi: 10.1080/02813432.2016.1207151. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Himmelreich JCL, Karregat EPM, Lucassen WAM, et al. Diagnostic Accuracy of a Smartphone-Operated, Single-Lead Electrocardiography Device for Detection of Rhythm and Conduction Abnormalities in Primary Care. Ann Fam Med. 2019;17(5):403–411. MLdoi: 10.1370/afm.2438. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Chalati MD, Shirvankar C, Rahimi S.. Machine learning models for atrial fibrillation detection in primary care using electronic health records: systematic review. Ann Fam Med. 2024;22(Supplement 1):6939. doi: 10.1370/afm.22.s1.6939. [DOI] [Google Scholar]
  • 25.DerSimonian R, Laird N.. Meta-analysis in clinical trials. Control Clin Trials. 1986;7(3):177–188. doi: 10.1016/0197-2456(86)90046-2. [DOI] [PubMed] [Google Scholar]
  • 26.Carter JL, Coletti RJ, Harris RP.. Quantifying and monitoring overdiagnosis in cancer screening: a systematic review of methods. BMJ. 2015;350(jan07 5):g7773–g7773. doi: 10.1136/bmj.g7773. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Johansson M, Zahl PH, Siersma V, et al. Benefits and harms of screening men for abdominal aortic aneurysm in Sweden: a registry-based cohort study. Lancet. 2018;391(10138):2441–2447. doi: 10.1016/S0140-6736(18)31031-6. [DOI] [PubMed] [Google Scholar]
  • 28.Ripping TM, Ten Haaf K, Verbeek ALM, et al. Quantifying overdiagnosis in cancer screening: a systematic review to evaluate the methodology. J Natl Cancer Inst. 2017;109(10). doi: 10.1093/jnci/djx060. [DOI] [PubMed] [Google Scholar]
  • 29.Voss T, Krag M, Martiny F, et al. Quantification of overdiagnosis in randomised trials of cancer screening: an overview and re-analysis of systematic reviews. Cancer Epidemiol. 2023;84:102352. Article 102352. 2023. doi: 10.1016/j.canep.2023.102352. [DOI] [PubMed] [Google Scholar]
  • 30.Bell K, Doust J, Sanders S, et al. A novel methodological framework was described for detecting and quantifying overdiagnosis. J Clin Epidemiol. 2022;148:146–159. doi: 10.1016/j.jclinepi.2022.04.022. [DOI] [PubMed] [Google Scholar]
  • 31.Lopes RD, Atlas SJ, Go AS, et al. Effect of screening for undiagnosed atrial fibrillation on stroke prevention. J Am Coll Cardiol. 2024;84(21):2073–2084. [DOI] [PubMed] [Google Scholar]
  • 32.Steinhubl SR, Waalen J, Edwards AM, et al. Effect of a home-based wearable continuous ECG monitoring patch on detection of undiagnosed atrial fibrillation: the mSToPS randomized clinical trial. JAMA. 2018;320(2):146–155. doi: 10.1001/jama.2018.8102. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Svennberg E, Friberg L, Frykman V, et al. Clinical outcomes in systematic screening for atrial fibrillation (STROKESTOP): a multicentre, parallel group, unmasked, randomised controlled trial. Lancet. 2021;398(10310):1498–1506. doi: 10.1016/S0140-6736(21)01637-8. [DOI] [PubMed] [Google Scholar]
  • 34.Halcox JPJ, Wareham K, Cardew A, et al. Assessment of remote heart rhythm sampling using the alivecor heart monitor to screen for atrial fibrillation: the REHEARSE-AF study. Circulation. 2017;136(19):1784–1794. doi: 10.1161/CIRCULATIONAHA.117.030583. [DOI] [PubMed] [Google Scholar]
  • 35.Gladstone DJ, Wachter R, Schmalstieg-Bahr K, et al. Screening for atrial fibrillation in the older population: a randomized clinical trial. JAMA Cardiol. 2021;6(5):558–567. doi: 10.1001/jamacardio.2021.0038. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Perez MV, Mahaffey KW, Hedlin H, et al. Large-scale assessment of a smartwatch to identify atrial fibrillation. N Engl J Med. 2019;381(20):1909–1917. doi: 10.1056/NEJMoa1901183. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Essén A, Stern AD, Haase CB, et al. Health app policy: international comparison of nine countries’ approaches. NPJ Digit Med. 2022;5(1):31. doi: 10.1038/s41746-022-00573-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.The Economist . Amazon makes a new push into health care. The Economist. 2022. https://www.economist.com/business/2022/11/20/amazonmakes-a-new-push-into-health-care. [Google Scholar]
  • 39.Chandrasekaran R, Katthula V, Moustakas E.. Patterns of use and key predictors for the use of wearable health care devices by US Adults: insights from a national survey. J Med Internet Res. 2020;22(10):e22443. doi: 10.2196/22443. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Kekade S, Hseieh C-H, Islam MM, et al. The usefulness and actual use of wearable devices among the elderly population. Comput Methods Programs Biomed. 2018;153:137–159. doi: 10.1016/j.cmpb.2017.10.008. [DOI] [PubMed] [Google Scholar]

Articles from Scandinavian Journal of Primary Health Care are provided here courtesy of Taylor & Francis

RESOURCES