gms | German Medical Science

Background: Digitalisation in healthcare should improve outcomes and increase efficiency of care. To this end, the amount of information that has to be dealt with is relevant but unknown. Up to now, complexity of drug therapy has been judged based on therapy at a single point of time. Data on drug therapy over a decade are lacking.

Materials and Methods: Database: BARMER health insurance data between 2011 and 2020 of 4.480.720 persons continuously insured in this period and aged 40 or older in 2020. Healthcare service utilisation was observed over a decade of life. Morbidity was measured by counting the number of the 241 ICD-10-GM diagnosis groups [1] that aggregate three-digit ICD-Codes into disease groups, e.g. E10 to E14 to “diabetes mellitus”. The number of doctor’s offices visited (distinct business location number (BSNR)), as well as the number of practices issuing a drug prescription were investigated. The number of prescription forms given to the patient in outpatient care were counted, as well as the number of prescribed drug packs, prescribed different active ingredients (7-digit ATC-Code) and prescribed ATC-code groups (3-digit ATC-code). Furthermore, the number of pharmacies visited to redeem prescriptions and the number of hospital stays with at least one overnight stay were evaluated.

Results: Between 2011 and 2020, on average, insured persons visited 21 different doctor’s offices and received prescriptions from seven different practices. Furthermore, an insured person had on average 37 diseases documented, received 76 prescriptions, 113 drug packs, and 20 different active ingredients of 11 ATC-code groups. The insured persons redeemed the prescriptions in on average six pharmacies and had two overnight stays in hospitals. The ten percent of insured persons with the highest utilisation of health services (90% percentile) visited at least 35 doctor’s offices, had 60 diseases documented and got 170 prescriptions, 270 drug packs, and 38 different active ingredients of 20 ATC-code groups prescribed by 13 physicians. The prescriptions were redeemed in at least 11 pharmacies and the insured persons had at least 6 overnight stays in hospital.

Conclusion: Over a decade, an average number of 20 drug ingredients were prescribed to a patient and an average number of 21 physicians treated them. This highlights the need for digitally supported documentation of drug therapy and for care coordination. Patients get their drugs on average from six up to 11 and more (upper 10% of patients) pharmacists. Neither a single physician nor a single pharmacist can oversee the therapy. It is unrealistic to assume that manual documentation of drug therapy is feasible. As manual documentation of prescribed drugs is extremely time consuming and can hardly be kept up to date, information on prescribed and dispensed drugs should be generated and stored in the healthcare process automatically – without effort for physicians, pharmacists and patients. Data have to be stored in a standardised and interoperative format to be usable for electronic decision support systems and for digital support of care coordination. Standardised interfaces have to be defined to enable the data exchange between inpatient and outpatient care and pharmacies. Without these preconditions, medication safety will not be achieved. Several projects funded by the G-BA Innovationsfonds (AdAM, TOP and eRIKA) apply this approach by using health insurance companies claims data and electronic decision support to inform physicians and to enable them to improve quality and safety of care [2].