Table 6.

Quality of the evidence, in aggregate, that supports antibiotic retreatment in patients with persistent symptoms of Lyme disease.

No. of studies	Limitations	Precision	Consistency	Indirectness	Evidence quality
4	Designated treatment effects were excessive [17,18] Unsupported design assumptions [17,18] Lack of pertinent inclusion criteria [17] Enrollment of treatment-refractory subjects	Small sample sizes (range 37–78) [16–18] Underpowered trial/end point [16,17]	Consistent finding of treatment effectiveness on fatigue in the trials by Krupp et al. and Fallon et al. [16,17]. Inconsistent findings on treatment effectiveness between the trials by Krupp et al., Fallon et al. and Klempner et al. [16–18].	Subjects had prolonged illnesses [16,18] Subjects had a history of extensive antibiotic treatment [16–18] Excluded subjects with co-morbidities and medication use commonly seen in practice [16–18] Restricted use of non-antibiotic medications, limiting practice [16–18]	Very low