Using routine data for quality of care assessments: A critical review, taking quality indicators for the "National Disease Management Guideline for Chronic Heart Failure'' as an example

In December 2009, the first version of the German Disease Management Guideline (DM-CPG) for chronic heart failure was completed, including a set of proposed quality indicators for heart failure. This article explores whether proposed indicators can be derived from data collected routinely in general practices. For this purpose, previous experiences and data from the research project CONTENT (CONTinuous morbidity registration Epidemiologic NeTwork) conducted under guidance of the Department of General Medicine and Health Services Research at the University of Heidelberg, Germany, were applied. The availability of numerators and denominators needed for calculating the four quality indicators for diagnosis and pharmacotherapy proposed in the DM-CPG was checked within so-called "routine data'' from the existing dataset of the CONTENT project. Within the given context, routine data are defined as data that are periodically transmitted from health care providers to cost units within the health care system. A thorough assessment has revealed that within the given context only one indicator could be deduced from routine data collection. This was the indicator measuring the proportion of patients receiving beta receptor antagonists, compared to all patients with heart failure NYHA class II to IV. Indeed, this single indicator will only be computable if the NYHA grade of heart failure severity and the presence or absence of contraindications to beta receptor antagonist therapy are routinely collected and the data merged into a central database. Against the background of these results it is obvious that a fully developed, transsectoral concept for data collection and data transfer needs to be implemented.