Identifying psychiatric patients' pathways of care by record linkage after pseudonymisation: linking inpatient and outpatient data for the total population of a province of Austria

Abstract

Background/Objectives: Health service utilization is usually described by event statistics, e.g. by frequencies or rates of ambulatory contacts or of hospital discharges for a given geographical area. Linking of patients’ records across health care domains, i.e. depicting typical pathways of patients, is not easily possible. However, such record linkage would be extremely important for health care planning and realistic cost calculation, for instance in order to identify typical referral patterns, repeated utilization of the same type of service („revolving door psychiatry“), or more than average utilization of the health care system („heavy users“). The objective of the study was to explore the possibility of linking inpatient and outpatient health service utilization data of psychiatric patients after discharge from hospital for the total population of the Austrian province of Lower Austria and to describe relevant pathways of care, including non-psychiatric services, such as general practitioners and non-psychiatric wards of general hospitals. Methods: Austria has an obligatory health insurance system with around 98% of the population insured, however with many different health insurers and patient records in many different data banks. Since a unique personal social insurance number is recorded whenever a health service is used, records can be principally linked in order to identify typical pathways of care. Because of ethical reasons, a record linkage approach after pseudonymisation of patient identification was employed. Among other problems the issue of different semantics in different data banks had to be solved. Results: 4,027 persons living in the Austrian province of Lower Austria (1.5 million inhabitants) were discharged from a psychiatric hospital bed in 2006 and were individually followed up for renewed utilization of health services within 12 months after discharge. 32.6% of these patients were readmitted to a psychiatric hospital bed (two thirds thereof already in the first three months). Even more, 36.4% were readmitted to a non-psychiatric hospital bed. After 3 months, 75% had seen a general practitioner and more than 85% had filled a prescription. Discussion/Conclusions: It proved feasible to link records by pseudonymisation in a fragmented health insurance system. Non-psychiatric service use is astonishingly high after discharge from a psychiatric hospital bed. Limitations of the study include the restricted possibility to identify some follow-up variables in a consistent way across the many health insurers. For instance, it was not yet possible to identify utilization of psychiatrists in an ambulatory setting in a consistent way. Also, by nature of the data available, utilization of social services, which is known to be quite high in psychiatric patients, could not be analysed. However, pathway data provide an important source of information about how a given health care system is functioning and invite further analyses explaining the identified pathway patterns. It is suggested that specific incentives and disincentives of provider payment system can explain some of the findings of pathway studies and that costs might be reduced by better information on heavy users. Funding: Austrian Ministry of Health (grant no. BMG-71603/0043-II/A/1/09). Keywords: Cost of illness, efficacy, pathways of care.