JMIR Formative Research · 4 authors, 2 centres
This summary was generated by AI from a single paper. It has not been reviewed by a clinician and is not clinical advice. Verify against the source before acting on it.
This randomized crossover study of 104 patients in rural Gujarat found 74% diagnostic concordance and 79.8% treatment concordance between provider-to-provider telemedicine and in-person care, with highest agreement for hypertension (95%) and diabetes (93%). Telemedicine is a safe and acceptable alternative when in-person care is not accessible, especially for managing high-burden primary care conditions in remote settings.
**Background:** Equitable access to primary health care remains a challenge in rural India, where 77% of qualified health workers are based in urban centers. Provider-to-provider telemedicine, connecting frontline health workers at primary care clinics with remote doctors, has been scaled up during the COVID-19 pandemic. However, rigorous randomized evaluations comparing telemedicine to in-person care in developing countries are scarce. This study aimed to compare diagnosis and treatment decisions from teleconsultations to those of in-person care in teleclinics in rural Gujarat.
**Methods:** A diagnostic concordance study using a randomized crossover design was conducted at 10 telemedicine primary care clinics in Morbi district, Gujarat. A total of 104 patients were enrolled and randomly assigned to first receive an in-person doctor consultation (59/104, 56.7%) or a health worker-assisted telemedicine consultation (45/104, 43.3%). The two groups were then switched. The in-person doctor and remote doctor were blinded to each other's diagnosis and management plan. The in-person doctor's diagnosis was considered the gold standard. The MyTeleDoc app, featuring a digital assistant called Ayu, guided community health officers (CHOs) in collecting patient history and physical exam information. Concordance was measured as the percentage of cases with the same diagnosis and treatment plan. Cohen kappa was calculated for hypertension and diabetes. Fisher exact tests were used to assess associations.
**Key Results:** Overall diagnostic concordance was 74% (77/104) and treatment concordance was 79.8% (83/104). No significant association was found between concordance and order of consultation (P=.65 for diagnosis, P=.81 for treatment), CHO-doctor pair (both P=.93), patient gender (both P>.99), or mode of teleconsultation (synchronous vs asynchronous; P=.32 and P=.29, respectively). A significant association was seen with type of case (P=.004 for diagnosis, P=.03 for treatment). The highest diagnostic concordance was for hypertension (20/21, 95%; Cohen kappa=0.93) and diabetes (14/15, 93%; Cohen kappa=0.89). The lowest concordance was for cardiology (1/3, 33%) and nonspecific symptoms (3/10, 30%). In 84% of cases, the doctor felt the information collected by the CHO via the digital assistant was sufficient to arrive at a diagnosis without needing to speak with the patient (asynchronous consultation). In 12 of 27 (44%) discordant cases, missing or incorrect information from the CHO contributed to discordance.
**Clinical Implications:** Telemedicine demonstrated acceptable diagnostic and treatment concordance with in-person care, particularly for common chronic conditions like hypertension and diabetes. The use of a digital assistant enabled more efficient consultations and adherence to evidence-based protocols. Telemedicine is a safe and effective alternative when in-person care is not accessible, especially in remote rural settings. It can improve health care access, reduce patient costs, and decrease overcrowding at tertiary facilities. However, limitations include lower concordance for conditions requiring physical examination (e.g., musculoskeletal, dermatology) and the potential to miss incidental findings. Training for CHOs and remote physicians on the limitations of telemedicine and indications for referral could further improve quality of care.