Comments on Brazilian workshop model for training investigators in chronic graft-versus-host disease according to the 2005-2006 National Institutes of Health recommendations

Abstract

The article by Rita de Cassia Barbosa da Silva Tavares(1) describes the results of a training course for professionals interested in applying the 2005-2006 National Institutes of Health (NIH) Consensus criteria to the scoring and response measurement of chronic graftversus-host disease (GvHD). Fifteen professionals and nine patients participated in the three and a half hour training session where clinicians were taught the consensus criteria and tried to apply them to actual patients. The results yielded some interesting observations. First, despite the familiarity of these physicians with the syndrome of chronic GvHD, it proved to be very difficult for them to apply the NIH criteria. The article identifies a number of confusing or controversial clinical situations that the participants felt were not being appropriately scored in the NIH recommendations. Second, the experience showed that training in the NIH criteria is important and that it must be more in depth and ideally, repeated over time, to reach a degree of reproducibility adequate for response assessment in clinical trials. Third, 20 minutes did not appear adequate to review the medical history, do the assessments and record the results, suggesting that additional time needs to be set aside, at least at the beginning when physicians are unfamiliar with the assessments. Another study by Mitchell et al. found that physicians took a median of 21 minutes to collect just the response and functional measurements,(2) but the Brazilian group captured a more comprehensive battery of instruments and included review of medical records in the assessment period. The difficulty encountered in this training program confirms that chronic GvHD scoring and response assessment can be very complicated, time-intensive and dependent on the observer. However, many worthwhile endeavors require training, practice and substantial effort to acquire expertise and chronic GvHD assessment is unfortunately one of these skills. Considerable practical experience caring for these patients does not automatically make application of the NIH consensus criteria intuitive or easy. The next step is to identify the best training methods, using measures of accuracy and reproducibility to define efficacy, and if possible, to develop transportable training materials. However, it may not be possible to use photos or standardized patients and each group studying chronic GvHD will need to implement training procedures to ensure adequate skills in doing the critical assessments. Clinical trials are only as good as the outcome data collected. It would be wonderful to have a biomarker, radiographic study or pathologic sample that could provide an objective measure of the severity or response of chronic GvHD. Until such a test is available, chronic GvHD will remain a clinical diagnosis that requires clinical assessments. The authors of this article are to be congratulated for conducting this training session and more importantly, reporting their experiences so that others may learn from them.