Table 1 Candidate quality indicators excluded in the first and/ or second expert workshop

 General aspects/ efficiency

Avoidable specialist visits

Proportion of personswith diabetes or hypertension unnecessarily transferred to a specialist for an uncomplicated cause (only considered diagnos: diabetes and hypertension). Estonia national guideline. Uncomplicated is based on an expert evaluation of all diagnosis codes.

The operationalization of the QI is based on ambulatory diagnoses (not available in Swiss health insurance claims data. Approximation using Pharmacy Cost Groups seems inappropriate.

 General aspects/ efficiency

Dispersion between providers

1) Ratio of primary care professionals (e.g. GPs) to specialists

2) Modified, modified continuity index (MMCI): This index focuses on the dispersion between providers and is based on the number of caretakers and number of visits only.

Redundancy: similar to the existing indicator (“Number of different primary care physicians consulted by an individual insured person”).

 General aspects/ efficiency

Emergency GP visits

1) Rate of emergency visits for adults

2) Number of emergency visits

Based on health insurance claims no differentiation between emergency and non-emergency possible.

 General aspects/ efficiency

Medication after hospital discharge

1) Proportion of persons with polymedication after vs. before hospitalization

2) Proportion of persons with PIM prescription after vs. before hospitalization

Limited influence of the GP on hospital medication prescription. Measures quality of hospital care/hospital processes.

 Laboratory testing

Vitamin B12 testing

Proportion of persons who received one or more vitamin B12 test

Information about the indication for vitamin B12 testing/ symptoms is missing in the claims data.

 Laboratory testing

Ferritin testing

Proportion of persons who received 1 or more ferritin test

Information about the indication for ferritin testing / symptoms is missing in the claims data.

 Laboratory testing

Complex lymphocyte panel

Proportion of persons who received a complex lymphocyte panel and a CD4 counts

Relatively small number of cases, therefore not suitable for large scale measurement.

 Screening

Colonoscopy

Proportion of persons who received a colonoscopy within 10-year interval

Building a 10-year cohort is not practical in Swiss health insurance claims database

 Imaging

Radiography

1) Proportion of persons with repeated radiographies with same indication/ localisation

2) Proportion of persons with preoperative chest radiography in absence of a clinical suspicion for intrathoracic pathology

It is not feasible to evaluate the appropriateness of radiography in the claims data. Not specific to primary care. Limited influence of the GP to influence the radiography process of other health care providers.

 Imaging

Ostodensitometry

Proportion of persons received repeated osteodensitometry

Relatively small number of cases, therefore not suitable for large scale measurement.. Clinical information missing.

 Geriatric care

NSAIDs (≥65 years and older)

Proportion of persons with NSAID prescriptions

Complex clinical situations/ multimorbidity limit alternative therapies. Individual case review is needed.

 Drug safety

Potentially inappropriate use of antibiotics

1) Proportion of persons with ≥1 antibiotic prescription

2) Proportion of women with ≥1 chinolone prescription

3) Proportion of women with ≥1 chinolone prescription who had no urine test

Interpretation without clinical information impossible. Very helpful for decision makring in clinical practice, but not appropriate for aggregated measurement.

 Drug safety

Drug interaction

Proportion of persons with selected adverse drug interactions based on DEGAM S1 list

Heterogeneity in definitions/ lack of broadly accepted list of medication combinations. Low practical relevance and relatively small number of cases, therefore not suitable for large scale measurement.