Loss to follow-up among patients diagnosed with spinal tuberculosis at a tertiary hospital in Western Cape Province, South Africa: A retrospective cohort study.

BACKGROUND: Patients diagnosed with spinal tuberculosis (TB) at a major tertiary hospital in Western Cape Province, South Africa, are required to attend regular follow-up at the hospital's outpatient spine clinic and to remain on TB treatment for at least 9 months. This follow-up and lengthy treatment is intended to allow for specialist monitoring of TB treatment response and early identification of secondary complications, and to reduce the risk of recurrence. However, little is known about adherence to these recommendations.

OBJECTIVES: The main objectives were to describe (i) loss to spine clinic follow-up (LTFU), and (ii) TB treatment duration among patients diagnosed with spinal TB at the hospital. Secondary objectives were to investigate (i) the association between LTFU and treatment duration, and (ii) factors associated with LTFU.

METHODS: This retrospective cohort study included 173 adults diagnosed with spinal TB between 2012 and 2015 and investigated follow-up within 2 years from diagnosis. Clinical, demographic and appointment data were obtained from hospital records and a dataset provided by the provincial Department of Health. LTFU was presented as frequency (%) and as a survival analysis. TB treatment duration was reported as frequency <9 months or ≥9 months, and the association between LTFU and <9 months of treatment was investigated using relative risk (RR) with 95% confidence intervals (CIs). Univariate associations between explanatory variables and LTFU were investigated using simple logistic regression analysis.

RESULTS: Patients had a median (interquartile range) age of 36 (29 - 48) years and included 98 females (57%) and 151 individuals (87%) residing <50 km from the hospital. Primary outcomes were that 129 patients (75%) were LTFU within 2 years of diagnosis and 45 (30%) completed <9 months of treatment. The RR of <9 months of treatment was 1.62 (95% CI 1.39 - 1.88) among those LTFU compared with those retained in follow-up. LTFU was not associated with any of the clinical or demographic variables investigated.

CONCLUSIONS: Three-quarters of the patients did not complete follow-up at the tertiary hospital spine clinic, and almost one in three received <9 months of TB treatment. Remaining in spine clinic follow-up was significantly associated with receiving at least the minimum duration of TB treatment. However, LTFU could not be predicted from routine clinical and demographic information and is likely to be related to factors not accounted for in the current analysis.