Assessment of lung function abnormalities in adult patients with tuberculosis in a high HIV-prevalent setting and the impact of a pulmonary rehabilitation intervention to improve lung function, functional capacity, and quality of life

Doctoral Thesis


Permanent link to this Item
Journal Title
Link to Journal
Journal ISSN
Volume Title
Background: Globally, tuberculosis (TB) continues to be a major health problem. In the most recent World Health Organisation (WHO) Global Tuberculosis Report of 2019, TB was ranked as the leading cause of death from an infectious disease ahead of the human immunodeficiency virus (HIV) and acquired immune-deficiency syndrome (AIDS). In the 2019 WHO Global Report on TB, there is little information relating to TB post-cure effects and management. Although there is evidence that successful completion of TB treatment does not equate to normal lung function, there is growing need for research, both during and after TB treatment, on the extent of lung function abnormalities and how these impact on the individual's quality of life (QoL). Pulmonary rehabilitation programmes may provide a continuum of care for individuals with TB to address both lung function abnormalities as well as positively impacting on QoL. Objectives: The present PhD thesis aimed to provide insight into the extent of pulmonary disease in individuals with pulmonary TB during and near completion of TB treatment as well as to establish whether provision of a pulmonary rehabilitation programme (PRP) could address the research gap. To achieve this, three linked studies were undertaken in the form of observational (prevalence) study (Study 1), a systematic review (Study 2), and a randomised control trial (Study 3). Study One: Observational Study Objectives: The overall aim of the present observational study was to ascertain the prevalence of lung function abnormalities in first time, drug sensitive individuals living with TB, with or without HIV coinfection, at near completion (at least four months) of TB treatment. The specific objectives were to determine: i) baseline clinical and socio-economic profile, ii) baseline information pertaining to the QoL outcome measures of EQ-5D-3L and the St George's Respiratory Questionnaire (SGRQ), iii) measure lung function parameters, iv) establish the proportion of participants with normal or abnormal (obstructive, restrictive, or mixed) lung function and the severity of these, v) whether a correlation of lung function abnormalities with chest x-ray (CXR) abnormalities exist, vi) establish whether a relationship exists between lung function and QoL measures, and vii) identify the predictors of lung function abnormality in individuals being treated for active TB. Methods: A cross-sectional observational study using a sample of convenience was conducted. Inclusion criteria included all adult male and females between the age of 18-80 years with confirmed (smear positive or by CXR) drug-susceptible TB who were receiving treatment, with or without HIV coinfection, for at least four months (16 weeks). ii Participants were excluded from Study 1 if they were adult patients who had had previous TB episodes, recent severe chest trauma (within the previous three months), a recent history of pneumonia, known atopic asthma, chronic bronchitis, emphysema, bronchiectasis prior to TB diagnosis, cardiac failure, or any other unrelated respiratory disease as reported in their medical folder. Participants completed two QoL questionnaires (EQ-5D-3L and SGRQ), a self-designed clinical research form to collect descriptive data, a six-minute walk test (6MWT), CXR, and spirometry once off. Results: The sample of 305 participants were predominantly male (n=168:55; 1%), had a median age of 36 years (IQR:28-43), and had median time of 19 weeks (IQR:18-22) on TB treatment. Overall, 32% of the sample presented with abnormal lung function (obstructive=11%, restrictive=15%, and mixed=6%). Only 2.2% of the total sample had two or more co-morbidities. There was no statistically significant difference (p=0.29) in distance covered by participants who had obstructive compared to restrictive lung function abnormality. After logistic regression analysis of clinical and sociodemographic variables (multi-variate), only being older (56–65 years old) and being obese were statistically significant (p=0.02 and p=0.04 respectively). When considering QoL, only the domain of mobility for the EQ-5D-3L questionnaire was statistically associated with abnormal lung function (p=0.02). Linear regression modelling for continuous variables of lung function (FEV1, FVC, FEV1/FVC and percentage predicted of FEV1, FVC, and FEV1/FVC) with SGRQ, 6MWD, and CXR scores yielded no predictor. Conclusion: Overall, 32% of participants presented with abnormal lung function, which is lower than comparator studies investigating lung function in TB populations. Quality of life measures for most participants was considered good at the time of assessment. Limitations to Study 1 related to the absence of normative data for a healthy population relating to lung function and 6MWD to compare the findings in this TB population. Recommendations for future research would be to establish normative data for these outcome measures. Regarding lung function testing, it is recommended that training of correct execution of the spirometry techniques is performed prior to assessment as the technique may be unfamiliar compared to the routine tests done at clinic visits for individuals receiving TB treatment. iii Study Two: Systematic Narrative Review A systematic review was conducted to establish the evidence of the impact of non-pharmacological intervention programmes (pulmonary rehabilitation) in the rehabilitation of individuals living with TB on lung function outcomes. Methods: MEDLINE via Pubmed, CENTRAL, CINAHL, PEDro, Web of Science, Scopus, Science Direct, and African Index Medicus, including Google Scholar were searched (from January 1995 to December 2016 with an updated search in November 2018) for randomised control trials, quasi-experimental and pre-post-test studies on PRPsfor adult individuals with TB specifically with lung function measures as primary outcome. Results: In total, 1 705 studies were obtained from the search. Once duplicate studies were removed, 1 220 studies remained. The titles and abstracts of these studies were screened resulting in 1 210 studies being excluded. Therefore, 10 studies were potentially eligible. Once the full-text articles were assessed, four studies met the inclusion criteria. Of the included studies, only one was a randomised control trial, two studies were single arm before and after studies, and one study was a prospective non-randomised open trial (two arms). In total, there were 178 participants in these studies, with sample sizes ranging from 10 to 67 participants. All four selected studies had both male and female participants; however, overall, male participants were the majority with 69% versus 21% females. The mean age across the studies was 70 years. No statistically significant difference (p>0.05) was found regarding lung function parameters and the PRPs. No meta-analysis could be performed as data could not be pooled due to the differences in study characteristics and outcome measures. Conclusion: This review was unable to support or negate the use of pulmonary rehabilitation for individuals with TB primarily due to the lack of well-designed and executed randomised control trials. The studies showed that no effect on FEV1 was demonstrated. The researchers recommended that future research investigates the extent of pulmonary sequelae in patients after completion of TB chemotherapy in large-scale studies. Long-term follow-up in those who have had TB without surgical intervention should be prioritised to see the extent of lung function disorders in this population, particularly in countries on the high-burden list for the disease. A further recommendation is that well executed randomised control trials that control for biases to investigate pulmonary rehabilitation in populations of individuals with TB should be prioritised as there is a need to develop an evidencebased continuum of care. iv Study Three: Randomised Controlled Trial Objectives: The overall objective of study three was to determine what the impact of a contextually relevant PRP would have on individuals living with TB, with or without HIV co-infection, on outcomes related to lung function, functional capacity, and QoL. Methods: A pilot randomised, single blinded, pre-test-post-test design was used. Inclusion criteria were all adult males and females between 18-65 years with TB confirmed by Gene Xpert, irrespective of number of TB episodes, HIV status, or having chronic obstructive pulmonary disease. Participants had to be within their first week of TB treatment. Participants with only extra-pulmonary TB, recent severe chest trauma (within the last three months), a recent history of pneumonia (within one month), known atopic asthma, cardiac failure, or any other unrelated respiratory disease as reported in their medical folders or who had defaulted on their treatment were excluded. In addition to this, if participants failed the pre-participation health screening and were non-ambulate due to paralysis or amputation, they were also excluded. Fifty-eight participants were randomised into a control group (CG) receiving only pharmacological therapy and the intervention group (IG) who received pulmonary rehabilitation in addition to pharmacological therapy. The PRP was held for 12 weeks and consisted of two weekly sessions with a duration of 45 minutes each, which was delivered at a community centre. Participants completed two QoL questionnaires (EQ-5D-3L and SGRQ), a self-designed clinical research form to collect descriptive data, a three-minute step test, and spirometry at three time points (enrolment, at six weeks, and at 12 weeks). T-tests were conducted to determine the difference between means of the CG and IG for lung function parameters, functional capacity, and QoL outcomes. Results: There were 29 participants in each group. Regarding sex, age, and number of co-morbidities the two groups were well matched. Regarding HIV status, the CG had more participants that were HIV positive (n=22) and on anti-retroviral therapy (n=11) than their IG counterparts (n=13 and n=5 respectively). Nearly half of the participants had a first time TB diagnosis, with the participants in the IG having reported more recurring TB incidences overall (n=16 vs. n=13). A t-test for difference between means showed no statistical significance for the CG and IG regarding FEV1, FVC, and FEV1/FVC ratio for absolute or percentage predicted values. Forty-three percent of participants in the total sample had normal lung function at baseline, with the remaining participants being classified as having either obstructive (26%), restrictive (21%), or mixed (10%) lung function. At baseline, 48% of participants in the CG had abnormal lung function compared to 67% in the IG. At six weeks there was no change in the CG regarding lung abnormalities. However, the IG only had 33% abnormal lung function at the same time point. v Although there was no statistical significance for any of the lung function categories, there was a 42% improvement in normal lung function at six weeks in the IG compared to the CG at baseline. The median baseline number of steps taken by the CG was 79 steps (IQR:42-134) compared to 117 steps by the IG (IQR:84-154). A t-test conducted to test the difference between means for the CG and IG was statistically significant for the step test (p=0.002) at six weeks for the IG, but not at 12 weeks (p=0.13). No correlation was found between the SGRQ (QoL parameter) and any lung function parameter (p>0.05) at 12 weeks. Conclusion: Although the changes in lung function, functional capacity, and QoL did not reach statistical significance at completion of the PRP for the IG, the continued improvement in all the outcomes for the IG from 0 weeks to 12 weeks holds potential clinical significance.