OBJECTIVE: To explore the feasibility of using existing data to measure the impact of surveillance by the Health Care Inspectorate (IGZ) on public health. DESIGN: Retrospective, descriptive. METHOD: We examined the magnitude of the health problems of suicide, pressure ulcers and medication errors before and after surveillance by the IGZ. To do this, we used data from the IGZ, external data files and data from the literature. In addition, we assessed definitions and possible underreporting and overreporting of the results. We then determined whether an estimate could be made of the effect of the surveillance on the magnitude of the health problem using a time series design. RESULTS: Medication errors were not sufficiently defined to measure an effect of surveillance. The prevalence of pressure ulcers and the incidence of suicide could be quantified reliably using data from the Inspectorate. A time series design showed that both health problems declined after surveillance by the Inspectorate. However, these trends could not be attributed to the surveillance since there was no insight into the external factors that might have played a role. CONCLUSION: In case of clearly defined health problems, the extent of the problem can be defined using IGZ data and trends can be observed using a time series design. However, establishing a causal relationship between supervision and observed trends requires an experimental research design, such as a prospective randomized study or a stepped wedge design.
OBJECTIVE: To explore the feasibility of using existing data to measure the impact of surveillance by the Health Care Inspectorate (IGZ) on public health. DESIGN: Retrospective, descriptive. METHOD: We examined the magnitude of the health problems of suicide, pressure ulcers and medication errors before and after surveillance by the IGZ. To do this, we used data from the IGZ, external data files and data from the literature. In addition, we assessed definitions and possible underreporting and overreporting of the results. We then determined whether an estimate could be made of the effect of the surveillance on the magnitude of the health problem using a time series design. RESULTS: Medication errors were not sufficiently defined to measure an effect of surveillance. The prevalence of pressure ulcers and the incidence of suicide could be quantified reliably using data from the Inspectorate. A time series design showed that both health problems declined after surveillance by the Inspectorate. However, these trends could not be attributed to the surveillance since there was no insight into the external factors that might have played a role. CONCLUSION: In case of clearly defined health problems, the extent of the problem can be defined using IGZ data and trends can be observed using a time series design. However, establishing a causal relationship between supervision and observed trends requires an experimental research design, such as a prospective randomized study or a stepped wedge design.
Authors: Einar Hovlid; Jan C Frich; Kieran Walshe; Roy M Nilsen; Hans Kristian Flaatten; Geir Sverre Braut; Jon Helgeland; Inger Lise Teig; Stig Harthug Journal: BMJ Open Date: 2017-09-05 Impact factor: 2.692