CoChat

Diagnostic Concordance of Telemedicine as Compared With Face-to-Face Care in Primary Health Care Clinics in Rural India: Randomized Crossover Trial

Open access
Verma, Neha, Buch, Bimal, Taralekar, Radha, Acharya, Soumyadipta · 2023 · JMIR Formative Research

Published version of a preprint

The DOI matches its Crossref record; 33 of the 34 references it cites were found in a registry or an index; 10 of the 51 sources its claims cite were checked for support.

Abstract

BACKGROUND: With the COVID-19 pandemic, there was an increase and scaling up of provider-to-provider telemedicine programs that connect frontline health providers such as nurses and community health workers at primary care clinics with remote doctors at tertiary facilities to facilitate consultations for rural patients. Considering this new trend of increasing use of telemedicine, this study was conducted to generate evidence for patients, health providers, and policymakers to compare if provider-to-provider telemedicine-based care is equivalent to in-person care and is safe and acceptable in terms of diagnostic and treatment standards. OBJECTIVE: This study aims to compare the diagnosis and treatment decisions from teleconsultations to those of in-person care in teleclinics in rural Gujarat. METHODS: We conducted a diagnostic concordance study using a randomized crossover study design with 104 patients at 10 telemedicine primary care clinics. Patients reporting to 10 telemedicine primary care clinics were randomly assigned to first receive an in-person doctor consultation (59/104, 56.7%) or to first receive a health worker-assisted telemedicine consultation (45/104, 43.3%). The 2 groups were then switched, with the first group undergoing a telemedicine consultation following the in-person consultation and the second group receiving an in-person consultation after the teleconsultation. The in-person doctor and remote doctor were blinded to the diagnosis and management plan of the other. The diagnosis and treatment plan of in-person doctors was considered the gold standard. RESULTS: We enrolled 104 patients reporting a range of primary health care issues into the study. We observed 74% (77/104) diagnostic concordance and 79.8% (83/104) concordance in the treatment plan between the in-person and remote doctors. No significant association was found between the diagnostic and treatment concordance and the order of the consultation (P=.65 and P=.81, respectively), the frontline health worker-doctor pair (both P=.93), the gender of the patient (both P>.99), or the mode of teleconsultation (synchronous vs asynchronous; P=.32 and P=.29, respectively), as evaluated using Fisher exact tests. A significant association was seen between the diagnostic and treatment concordance and the type of case (P=.004 and P=.03, respectively). The highest diagnostic concordance was seen in the management of hypertension (20/21, 95% concordance; Cohen kappa=0.93) and diabetes (14/15, 93% concordance; Cohen kappa=0.89). The lowest values were seen in cardiology (1/3, 33%) and patients presenting with nonspecific symptoms (3/10, 30%). The use of a digital assistant to facilitate the consultation resulted in increased adherence to evidence-based care protocols. CONCLUSIONS: The findings reflect that telemedicine can be a safe and acceptable alternative mode of care especially in remote rural settings when in-person care is not accessible. Telemedicine has advantages. for the potential gains for improved health care-seeking behavior for patients, reduced costs for the patient, and improved health system efficiency by reducing overcrowding at tertiary health facilities.

Overview

CoChat Summary

TL;DR

In a randomized crossover study of 104 patients at 10 rural primary care clinics in Gujarat, telemedicine showed 74% diagnostic concordance and 79.8% treatment concordance versus face-to-face care. The authors conclude that provider-to-provider telemedicine, supported by a digital assistant, can be a feasible and safe alternative when in-person care is hard to access, especially for common conditions like hypertension and diabetes.

Key findings

  • Among 104 patients, diagnostic concordance between in-person and remote doctors was 74% (77/104), and treatment-plan concordance was 79.8% (83/104).
  • Concordance did not differ significantly by consultation order (P=.65 for diagnosis; P=.81 for treatment), health worker–doctor pair (both P=.93), patient gender (both P>.99), or synchronous vs asynchronous teleconsultation (P=.32 and P=.29).
  • Concordance differed by case type (P=.004 for diagnosis; P=.03 for treatment). Highest diagnostic concordance was for hypertension: 20/21 (95%), Cohen kappa=0.93 and diabetes: 14/15 (93%), Cohen kappa=0.89; lowest was for cardiology: 1/3 (33%) and miscellaneous/nonspecific symptoms: 3/10 (30%).
  • The digital assistant enabled asynchronous care in many cases: in 84% of cases, doctors felt the information collected by the community health officer was sufficient to diagnose without speaking to the patient.

Methods

  • Randomized crossover diagnostic concordance study in 10 telemedicine primary care clinics in rural Morbi district, Gujarat; 104 patients were enrolled after one excluded case.
  • Patients were randomized to receive either in-person first then telemedicine (59/104) or telemedicine first then in-person (45/104), with a 30- to 90-minute gap; the in-person diagnosis/treatment plan was treated as the gold standard.
  • Main analysis compared concordance between the two modalities using percent agreement, Fisher exact tests, and Cohen kappa for common diagnoses; analyses were done in SPSS and Excel.

Limitations

  • The paper states that it measured concordance, not accuracy, because there was no reference standard; it therefore cannot determine which modality was “correct” when diagnoses differed.
  • The authors note observer bias because providers spent more time than usual on consultations, though they think this likely affected both arms similarly.
  • The text makes evident that the study had small subgroup sizes for some conditions (for example, cardiology n=3), limiting conclusions for less common diagnoses; it also notes that 6.7% (7/104) of patients may have shared different complaints across consultations, which could have affected agreement.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.2196/42775

Cites

33 of 33 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Figure 1 Location of Morbi district. (reproduced from Milenioscuro [23], which is published under Creative Commons Attribution 4.0 International License [24]).
Figure 1 Location of Morbi district. (reproduced from Milenioscuro [23], which is published under Creative Commons Attribution 4.0 International License [24]).CC BY · PubMed Central
Figure 2 MyTeleDoc application with the digital assistant Ayu.
Figure 2 MyTeleDoc application with the digital assistant Ayu.CC BY · PubMed Central
Figure 3 A community health officer using the digital assistant Ayu in the MyTeleDoc app to collect a patient history and facilitate a teleconsultation.
Figure 3 A community health officer using the digital assistant Ayu in the MyTeleDoc app to collect a patient history and facilitate a teleconsultation.CC BY · PubMed Central

Verification

Claims the paper makes

8 Valid · 0 Not valid · 24 Not checked · 123 Own results and interpretations

60 more claims

Create a free account to see all 155 claims and every source verdict.

See all claims — free account
  • Probably not a substantial claim. 60
  • It may state two separate claims as one. 49
  • The quote only partly states this claim. 1

Create a free account to see these claims.

Citations

Cites

Show all 33 references

Metadata

Authors
Verma, N., Buch, B., Taralekar, R., & Acharya, S.
Affiliations
  • Baltimore Museum of Art (US)
  • Johns Hopkins University (US)
Corresponding author
Verma, Neha
Venue
JMIR Formative Research
Published
2023-06-23
Volume
7
Pages
e42775
ISSN
2561-326X (ISSN-L)
OpenAlex
W4365131587
Cited by
30
References
34
Citations by year
2026: 12 · 2025: 5 · 2024: 9 · 2023: 4
FWCI
7.65
Type
journal article
Language
en
Open access
Licence
cc-by
Locations
  • JMIR Formative Research (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Topics
  • Health Sciences › Medicine › Public Health, Environmental and Occupational Health › Telemedicine and Telehealth Implementation
  • Health Sciences › Medicine › Public Health, Environmental and Occupational Health › Healthcare Systems and Technology
  • Health Sciences › Medicine › Family Practice › Primary Care and Health Outcomes
Keywords
telemedicine, diagnostic agreement, primary care, rural health care, community health workers, antihypertensive therapy, diabetes management, chronic kidney disease, COVID-19 pandemic, Gujarat, patient satisfaction, health worker satisfaction

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

The claim verdicts this report counts are experimental: their accuracy is not yet measured.

A language model drew the claims from the paper on 2026-09-30, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-09-30 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (gpt-6-sol).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.