Answers are grounded in this paper and cite their sources.
Dear Editor, Eczema herpeticum (EH) is characterized by viral infection of a pre‐existing inflammatory dermatosis. EH is associated with considerable morbidity and even mortality. It is important to study EH in order to identify risk factors and burden of disease. However, the epidemiology of EH has not been well defined. To our knowledge, large‐scale population‐based studies that examined the epidemiology of EH have not been reported. Previous studies of risk factors for EH have been mostly limited to single centres and small cohorts. Self‐reported history of EH using survey‐based research is likely not a feasible epidemiological approach given the low incidence of disease. Evaluation of large‐scale clinical datasets that are representative of the broader population is needed in order to capture meaningful epidemiological data from a sufficient number of EH cases. To date, no reports have established a best method for detecting patients with EH from such datasets. The use of International Classification of Diseases, Ninth Revision, Clinical Modification (ICD‐9‐CM) diagnostic codes is a potential methodology for the detection of EH, though validity for this methodology has yet to be established. We recently validated the ICD‐9‐CM codes for pemphigus and pemphigoid and identified several strategies for improving the positive predictive value of such diagnostic codes.1 Previous studies have also validated ICD‐9‐CM coding for detection of other dermatological disorders, such as psoriasis2 and Stevens–Johnson syndrome/toxic epidermal necrolysis.3 We sought to determine whether the ICD‐9‐CM code of 054.0 similarly demonstrates a high predictive value for the confirmed diagnosis of EH in adults. The study protocol was approved by the Northwestern University Institutional Review Board. Electronic medical record (EMR) data mining was conducted through the Northwestern Medicine Electronic Data Warehouse, which includes an EMR repository for > 2 500 000 individuals from multiple affiliated Chicago metropolitan area sites. Data from individual adult patients who had a documented diagnosis of EH between January 2001 and November 2014 were included in this study based on the ICD‐9‐CM code 054.0. Case data, including free text, laboratory values and all unstructured data, were reviewed in their entirety from first encounter and included all subsequent outpatient and inpatient encounters. The diagnosis of EH was confirmed by the presence of a clinical diagnosis by a board‐certified dermatologist along with laboratory confirmation, including positive viral culture, Tzanck smear, direct fluorescence antigen, polymerase chain reaction and/or serology for herpes simplex virus (HSV). An EH rating scale with scoring from 0 to 3 was utilized to categorize an EH diagnosis based on clinical and laboratory findings. One point was given for clinical diagnosis, prior episode of EH or cold sores at time of episode, and at least one positive laboratory test consistent with EH. A half‐point was assigned if the diagnosis was included in the clinical or diagnostic differential diagnosis but the findings were not definitive and if the patient had a prior history of cold sores or other HSV infection. A score > 1 was considered to be confirmatory for EH and compared with the presence of a single occurrence or multiple occurrences of the ICD‐9‐CM code 054.0. Sensitivity, specificity, positive and negative predictive value and kappa coefficients were estimated. Overall, 64 patients with the ICD‐9‐CM diagnostic code 054.0 were identified. The mean ± SD ages at latest follow‐up of patients was 45·1 ± 17·0 years; 41 patients (64%) were female. We performed a manual review of the EMR and stratified results by frequency of use of the diagnostic code of interest (single, multiple) (Table 1). In outpatients who had a single occurrence of the diagnostic code 054.0 (n =43), 38 patients (88%) had no clinical or diagnostic documentation to support a diagnosis of EH (score = 0). In outpatients who had multiple occurrences of the diagnostic code 054.0, all patients had clinical and diagnostic information to support a diagnosis of EH (score > 1). The presence of multiple diagnostic codes of 054.0 was associated with excellent specificity and positive predictive value. Among the 13 patients who had an inpatient diagnostic code of 054.0, all had clinical and diagnostic information to support a diagnosis of EH (score > 1). Confirmation of International Classification of Diseases, Ninth Revision, Clinical Modification (ICD‐9‐CM) code 054.0 [eczema herpeticum (EH)] NPV, negative predictive value; PPV, positive predictive value. Laboratory confirmation includes positive viral culture, Tzanck smear, direct fluorescence antigen, polymerase chain reaction and/or serology for herpes simplex virus (HSV). Values are provided as n (%) unless otherwise stated. Confirmation of International Classification of Diseases, Ninth Revision, Clinical Modification (ICD‐9‐CM) code 054.0 [eczema herpeticum (EH)] NPV, negative predictive value; PPV, positive predictive value. Laboratory confirmation includes positive viral culture, Tzanck smear, direct fluorescence antigen, polymerase chain reaction and/or serology for herpes simplex virus (HSV). Values are provided as n (%) unless otherwise stated. There are multiple factors that contributed to erroneous ICD‐9‐CM coding of EH (Table 2). Firstly, EH can often be a challenging clinical diagnosis, even for experienced dermatologists and laboratory verification is often needed for confirmation, e.g. direct fluorescence antigen. Although a single ICD‐9‐CM code of 054.0 might be used when such laboratory testing is necessary, the final diagnosis is subject to change if subsequent laboratory testing is negative. In addition, typographical or ‘human error’ appears to have occurred in some cases due to similar sounding names of diagnoses, e.g. eczema and ICD‐9‐CM codes similar to other types of herpetic infections. These issues can be remedied by excluding patients with only a single occurrence of the diagnostic code for EH. Finally, some patients were diagnosed outside the institution and presented with an established history of EH. This introduces the possibility of misclassification bias and incomplete documentation of diagnostic testing at other institutions. Future quality improvement studies are warranted to reduce miscoding in EH. Scenarios of misuse of International Classification of Diseases, Ninth Revision, Clinical Modification code 054.0 [eczema herpeticum (EH)] HSV, herpes simplex virus. Data are presented as n (%). Scenarios of misuse of International Classification of Diseases, Ninth Revision, Clinical Modification code 054.0 [eczema herpeticum (EH)] HSV, herpes simplex virus. Data are presented as n (%). One strength of this study is that it includes a structured data abstraction to confirm the diagnosis of EH based on both clinical and laboratory findings. Another strength is that both outpatient and inpatient records were reviewed within a large quaternary care hospital and affiliated sites in a major metropolitan city. A limitation of this study is that the patient population may not be representative of the general population. However, if a single occurrence of an EH diagnostic code is not reliable in this setting, it is not likely to be reliable in most other settings. In conclusion, the results of this study do not support the hypothesis that a single occurrence of an ICD‐9‐CM code of 054.0 is sufficient for identification of EH in the outpatient setting. Excluding patients with only a single occurrence of the ICD‐9‐CM code 054.0 or limiting reviews to inpatient records appear to be better approaches. Moreover, the adoption of an EH scoring tool or equivalent approach to confirm an EH diagnosis may be warranted. Northwestern Medicine Enterprise Data Warehouse is supported by Northwestern University Clinical and Translational Sciences grant 8UL1TR000150.
No takes yet. Share an insight, caveat, or question.
Hsu et al. (2016) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: