[Internal diagnostic validation of patients with a chronic wound: possibilities of identification on the basis of...

There are still few epidemiological data on patients with chronic wounds (leg ulcers, diabetic foot ulcers and pressure ulcers). Statutory health insurance (SHI) data is increasingly being used for questions relating to healthcare science. When using this data, which is primarily collected for billing purposes, the methodological procedure for defining cases must be presented transparently. Here, it must be checked whether the target group can be validly defined using the coded diagnoses and, if necessary, further information from routine data. Therefore, the aim of this contribution is, on the one hand, to develop criteria with the help of which patients with a florid (active) chronic wound can be identified safely or as doubtful cases in routine data and, on the other hand, to determine the corresponding frequency estimates.

Initially, a literature research was carried out to identify parameters relevant to care in patients with chronic wounds. In the next step, these were divided into specific, less specific and non-specific criteria (visual validity) in a multi-stage consensus procedure with regard to the specificity for wound care. On this basis, three different case definitions are used to identify florid chronic wounds. Based on an SHI sample of insured persons, frequency estimates were made for various case definitions (safe and questionable cases).

Of the 21 parameters identified in the literature, eight were classified as specific, six as less specific and eight as non-specific criteria for the identification of patients with chronic florid wounds. Using diagnostic coding alone for the target diseases, an administrative prevalence of chronic wounds of 1.13% was observed for the year 2010. If a case is defined using the less specific and/or the specific criteria, prevalence drops to 0.79%; if only the specific criteria are used, prevalence drops only marginally to 0.78%. These changes were observed in patients with leg ulcers and diabetic foot ulcers, but not in patients with pressure ulcers. Here, the lowest administrative prevalence (0.18%) can be seen when only looking at the diagnoses, but this increases slightly when taking wound-relevant treatments into account (specific and less specific criteria: 0.25%).

It is possible to define patients with a chronic florid wound on the basis of wound-relevant treatments using SHI data and to make estimates of administrative prevalence. Depending on the question, the criteria for defining cases can be narrowed down or broadened. The comparison provides information on the internal validity of diagnostic coding. However, further studies are needed to verify external validity.