Response to meta-analysis of Lyme borreliosis symptoms.

Cairns and Godwin 4 in their meta-analysis of five selected United States studies 5–9 on the outcome of Lyme borreliosis concluded that the prevalence of the symptoms of persistent fatigue, musculoskeletal pains or neurocognitive difficulties (post-Lyme disease syndrome, chronic post-treatment Lyme disease, or chronic Lyme disease) exceeds that of comparison populations without Lyme disease by >5%. Based on these findings, a conservative estimate of the incidence of this syndrome in the United States would be at least 3000 cases per year (>20 000 reported cases ·3; to include unreported cases, ·0.05). Given the chronic nature of the process, the prevalence of cases should be many times higher. Furthermore, post-Lyme disease symptoms are occurring on a high background rate of indistinguishable symptoms in the general population. Based on the studies cited in the meta-analysis, the frequency of musculoskeletal symptoms alone among the control groups of adults was 25.8%. Thus, more than 30%of adults after treatment for Lyme disease (25.8% 1 .5%) could be regarded as having post-Lyme disease syndrome. This prompts one to wonder why so many difficulties were encountered in enrolling patients in four different controlled treatment trials of Lyme disease patients with post-treatment symptoms. 12–14 Three of the four studies never met their enrollment goals, and the only study 13 to meet its enrollment objective of 55 patients included several patients, that, in retrospect, failed to satisfy entry criteria. Taken together, all four studies managed to recruit ,200 patients during an enrolment period of several years. The findings of this meta-analysis 4 are uninformative and misleading, since the studies upon which it was based have substantive limitations including retrospective study design, lack of timely antimicrobial therapy (or in some instances no treatment at all), use of antimicrobial regimens no longer recommended, 15 lack of 2-stage conditional serologic testing to establish the original diagnosis of Lyme disease, 16 a distorted case mix with an over representation of patients with extracutaneous manifestations, failure to consider pre-Lyme disease traumatic psychologic experiences, 17 and the likely inclusion of an unknown number of non-Lyme disease patients whose original complaints were exclusively subjective in nature and consisted of chronic arthralgias, fatigue, or cognitive difficulties. The last group of patients may have been evaluated for Lyme disease based on the mistaken notion that patients with such complaints are likely to have ‘chronic Lyme disease.’ What is more and very important is that the laboratory testing in many of the studies used older serological methods. Methods that are now known to be associated with high rates of false positive results 18 are no longer considered valid. Indeed, the majority of patients who carry the diagnosis of ‘chronic Lyme disease’ have no evidence of ever having been infected with Borrelia burgdorferi. 19 It should be noted that in the study included in the meta-analysis involving the largest number of subjects, 5 at least one-third did not meet the national surveillance case definition for Lyme disease. In many studies of ‘chronic’ Lyme disease a form of ‘protopathic bias’ can be a major problem. 20 Such bias can occur when persons are diagnosed with ‘chronic Lyme disease’ because they have the outcomes of interest (e.g. chronic pain, fatigue). In addition, it is likely that there is reporting or recall bias in patients who are labeled as having Lyme disease. 21 Thus, a person diagnosed with Lyme disease would be more likely to recall and/or to report subsequent symptoms such as arthralgia, myalgia or fatigue than would another person with the same symptoms who was never diagnosed as having Lyme disease. In this context, the Seltzer study 5 showed that responses to questions about the ability to conduct the normal tasks of daily living were not significantly different than those of controls without Lyme disease. In addition, the results of a battery of formal neuropsychological tests administered to patients who had Lyme disease were normal in the study reported by Vazquez et al. 8