Khalid, Tengku Noor Farhana Tengku
(2022)
Development and effectiveness of tuberculosis risk score for smokers in tuberculosis screening uptake and detection.
Masters thesis, Universiti Sains Malaysia.
Abstract
Background: Tuberculosis (TB) remained one of the major infectious diseases
worldwide. Smokers are considered as high-risk for TB disease. Many TB cases are
still undetected, especially among the smoker population.
Objective: To determine the factors associated with TB diseases among the smoker
population, developing, validating, and evaluating the effectiveness of the
Tuberculosis Risk Score for Smoker (TBRSS) in Bachok, Kelantan.
Method: This study consists of three phases. The case-control study in phase 1
consists of 159 patients comparing patients with TB and those without the disease.
Their smoking status was either active, ex- or passive smokers aged 18 years and
above. The patients were interviewed using proforma (45 items). Simple and
multiple logistic regressions were applied to analyse the data. In Phase 2, the risk
score was derived from multiple logistic regression or decision tree classification
analysis. The content validation study involved eight TB experts. Values were
obtained for factor content-validity index (F-CVI) and scale-level content validity
index, averaging index (S-CVI/Ave). Twenty healthcare workers (HCWs) in health
clinics were selected for the face validation study and face validity index (FVI)
values were obtained. The quasi-experimental study in phase 3 involved 156
participants 18 years and above known for their smoking status screened with
TBRSS. The proportion of patients who underwent screening and diagnostic tests at
different dates was noted. The number of smokers screened before and after
implementing the TBRSS was noted. The difference between the regression line and
R2 before and after were obtained based on interrupted time series analysis.
Results: In the Phase 1 study, most participants are male with a median age of more
than 40 years. The patients consisted of a mixed number of active, ex- and passive
smokers. Almost all patients never had TB before. More than half of the patients had
cough with sputum, did not have night sweat, no chest pain and no haemoptysis. The
median duration of TB symptoms is 30 days for the cases and seven days for the
controls. Four variables were significantly associated with TB disease: smoking
status, presence of night sweat, significant weight loss and duration of TB symptoms.
From that, two risk scores developed. The experts gave the F-CVI and S-CVI/Ave
values of 0.83 and above, and the HCWs gave the FVI values of equal or more than
0.88. The TB experts chose the risk score derived from multiple logistic regression as
the TBRSS. In the Phase 3 study, the accuracy of the TBRSS is 20.5%, with a
sensitivity value of 80.0% and specificity value of 18.5%. The positive predictive
value is 3.1% and the negative predictive value is 96.6%. Most participants (87.2%)
had their diagnostic tests done on the same day with screening. The regression line
improved with the R2 value changing from 0.064 to 0.160 during the intervention
Conclusion: The TBRSS is a validated newly developed TB screening tool. It
effectively increases the number of TB screenings among the smoker population.
More studies need to be conducted to evaluate the usefulness of the TBRSS further.
Actions (login required)
 |
View Item |
