[Outpatient diagnoses in primary care: Adding transparency in the field of outpatient diagnoses]

In Switzerland, there is no obligation to record outpatient diagnoses. PraxisGruppe Schweiz (PraxiS), a network of about 30 GP and specialist practices in predominantly rural German-speaking regions in Switzerland, assign diagnosis codes per patient and consultation using the ICD-10 classification system. We compare these diagnoses with approximate ones, the so-called Pharmaceutical Cost Groups (PCGs), based on claims data from SWICA health insurance. We suspected that the distributions of diagnoses differed significantly between the two samples.

The PraxiS database was used to identify patients with at least one consultation between 2020 and 2022 who were insured with SWICA. Claims data identified patients who had at least one consultation with a PraxiS network clinician in the same period. Diagnoses (ICD-10) were analyzed, and the ten most frequent diagnoses per age group were reported. The number of consultations per patient and age group were calculated. PCG information was derived from claims data following the official Swiss risk equalization definition, based on defined daily doses in the last observation year. Differences in diagnosis distributions between the two samples were tested using a Pearson χ² test.

High cholesterol, hypertension, diabetes, and depression are amongst the most frequently recorded diagnoses in both data sources. Furthermore, acute events (e.g., infections of the upper respiratory tract) and non-specific symptoms or problems that cannot be clearly assigned to a certain diagnosis occur frequently in primary care, while thyroid diseases, asthma/COPD and glaucoma are more commonly recorded in the claims data, since they are associated with regular medication intake. For 6 of the 32 PCG diagnoses, we rejected the null hypothesis of identical distributions, i.e., the diagnosis prevalences differ significantly between the two samples.

While the GPs' diagnoses reflect the most frequent reasons for consultations, claims data also include prescriptions from outside primary care. Both data sources must be used in a complementary manner. Claims data should be used to analyze morbidity and medication use in a population. To gain a deeper understanding of the care situation, the data generated by GPs should be used. Particularly in times of GP shortages, when it is necessary to think about alternative or advanced forms of primary care, a detailed analysis of the care needs and care situations of the affected patient groups is necessary. Efforts in this direction should be supported.