A four-year trend in pulmonary bacteriologically confirmed tuberculosis case detection in Kampala-Uganda

Tuberculosis (TB) remains a global public health problem despite the presence of TB pharmacotherapy for more than 50 years and the use of vaccines for more than 90 years [1, 2]. According to the World Health Organization (WHO), TB remains one of the major global health threats of the twenty-first century [3] as well as a major cause of socio-economic distress [4, 5]. According to the 2016 Global Tuberculosis (TB) Report, in 2016 alone an estimated 10.4 million people developed TB and 1.4 million died from the disease with over 95% of these deaths occurring in low- and middle-income countries. An estimated 11% of the incident TB cases in 2015 were concurrently human immunodeficiency virus (HIV) positive [6].

TB has remained a public health challenge in low resource settings and, Uganda, ranked as one of the highest TB and TB/HIV burden countries in the world, remains disproportionately affected [6]. Uganda's TB prevalence is estimated at 253 per 100, 000 population [6]. Since 1998, the Uganda National Tuberculosis and Leprosy Program (NTLP) has implemented a series of WHO TB control strategies. The first of which was the Directly Observed Treatment Short Course (DOTS) strategy, which allows patients to take their daily drugs under the observation of health professionals, thereby improving treatment compliance and increasing the cure rate [7, 8]. This was replaced by the more comprehensive Stop TB Strategy implemented in 2006, followed by the End TB Strategy in 2015, to address the emerging challenges of drug-resistant TB, TB/HIV co-infection and the incorporation of new diagnostic tests, such as GeneXpert [9].

Reports have shown that PCB TB patients are more likely to transmit TB to susceptible hosts compared to those with a bacteriologically negative status [10]. A high proportion of PBC TB cases might imply more reliance on bacteriological examination to diagnose TB or a delay in diagnosis of both extrapulmonary (EP) and clinical pulmonary (PCD) TB which with time become easier to diagnose bacteriologically. On the other hand, low proportions may reflect several gaps in diagnosis such as lack of capacity by the program to accurately diagnose TB through bacteriological examination. Such programs majorly rely on clinical diagnosis basing on the presence of radiographic abnormalities on the radiograph or based on a high index of clinical suspicion which carries along with risks of falsely treating patients with other conditions for tuberculosis [11].

Although the proportion contributed by PBC TB patients to the total case notification is important in the monitoring of the quality of the national program, the trend of their notification has not been carefully studied to further inform the epidemiology of TB in this setting. We sought to determine the trend and outcomes of PBC TB over the four-year period of January 2012 –December 2015 in Kampala City.

We assessed a total of 10,404 records of TB patients from the 65 active DTUs in the five divisions of Kampala during the period of January 2012 to December 2015. Of the 10,404 records of TB patients; 6551 (63.0%) were PBC, 2207 (21.2%) were pulmonary clinically diagnosed (PCD) and 1343 (12.9%) were extrapulmonary (EP) while 2.9% had no documentated classification of TB in the register.

The PBC TB patients had a mean age of 31.92 (SD ±11.24) years. The majority of PBC TB patients were male, 3734(57.0%). PBC TB patients contribution by division from highest to lowest was Kawempe, Lubaga, Makindye, Nakawa, and Central respectively, (see Table 1 for details). About 74.9% of all PBC TB patients were diagnosed by sputum smear microscopy test, 15.6% were diagnosed by GeneXpert testing, and only four patients (0.1%) were diagnosed by culture testing. The remaining records were undocumented and classified as unknown.

This study analyzed TB patient records from January 2012 to December 2015 in all registered DTUs of the most highly burden district in Uganda. Our discussion of PBC TB patients gives detail that according to our knowledge has not been examined before in this country and district. As widely observed throughout the world and in the country, most of the PBC TB patients were male [14‐16] in their young and productive years [15]. This is emphasized by the clear-cut gender differential that has been observed in many other studies [14, 16, 17] and the economic impact of TB due to its high incidence among people in the productive years [17].

Majority of the PBC patients were from Kawempe division, this could be due to the location of Mulago National Referral Hospital, the only national referral hopsital in the country, which also contributes to the highest number of TB cases notified each year in the district. About 2% of the PBC TB patients had an unknown/undocumented HIV status and it was noted that less than half of all PBC TB patients were HIV positive patients, a finding that was not surprising since it has been earlier noted by researchers that HIV positive TB patients are most likely to show up with EP TB [18].

The statistically significant increase in PBC TB patients diagnosed across the period for this study could be due to improvement in the quality of microscopy done leading to a better performance in the diagnosis of TB using smear microscopy. This could further be attributed to the NTLP’s adoption of the use of GeneXpert machines to diagnose TB where 15.6% of all PBC patients were diagnosed by GeneXpert. The program and its partners have installed over 13 GeneXpert machines within the district and released new guidelines in 2017 for the adoption of GeneXpert as the first diagnosis method for TB [19]. Given that GeneXpert has a higher sensitivity than microscopy [20], this could account for the registered increase in PBC TB pattients. It is to the knowledge of the authors of this paper that the NTLP moved forward to classify all patients that were GeneXpert positive as PBC patients [21] despite the fact that GeneXpert machines were in use before this decision was made. This may have affected the number of patients classified as PBC for a short period. It is worth noting that there was a significant increase in all classifications of TB within the study period.

We observed that through the study period, patients diagnosed with a 3+ (>10AFB) were persistently higher compared to those that were 2+ or 1+. The authors speculate that this could be caused by late diagnosis due to patient delay in seeking care, leading to an increase in the severity of infection, thus leading to more bacilli detection per length through microscopy. This possibility of total delay in seeking health care and diagnosis may contribute to unfavorable treatment outcomes [22]. In addition, a late diagnosis has implications on cross-transmission of TB to the rest of the population, health workers and other patients at the health centers [23]. However, the same results when expressed as proportions for each year revealed an increase in patients being diagnosed 1+ compared to 2+ and 3+ by the year 2013 that may be attributed to the initiation of active case finding/TB screening using community linkage facilitators (CLFs) leading to early case detection. The contribution of active case finding/TB screening may further be realized by the general decline in proportions of 3+ PBC TB cases diagnosed through the years. This may also be explained by the fact that there were less severely sick patients reporting to the facilities making it less likely diagnose patients as 3+ PBC TB compared to the baseline year 2012 where more severely sick patients would easily be diagnosed. Therefore, this observation could be attributed to both the use of active case finding and use of CLFs.

In Kampala, engagement of the private health provider is important in TB control. Most patients are commonly known to first seek care through private health care providers. There was a decline in PBC TB diagnoses by private facilities in 2015 that this study did not explore further. Researchers, however, speculate that this could be due to the commencement of referral of specimen from private facilities to public facilities for GeneXpert testing. Our research showed a statistically significant increase in GeneXpert testing through the years and this could have influenced the increasing trend of PBC TB patients observed during our study period. This also proved that smear microscopy alone could lead to underestimation of the burden of smear-positive pulmonary TB. More sensitive testing methods such as GeneXpert could provide a better estimate.

Over the years, there was a statistically significant increase in the number of TB/HIV co-infected PBC patients that were linked to HIV care. The proportion of PBC TB/HIV co-infected patients on CPT and ART improved from 37.7% to 51.9% and from 35.5% to 48.4% respectively. This observation could be accounted for by the emphasis on TB/HIV collaborative services by the NTLP and its partners. However, these results still reflected a low uptake of both CPT and ART. This may also be attributed to poor record keeping since many of the PBC TB/HIV co-infected patients did not have a documented status of their CPT and ART uptake.

There was an overall statistically significant improvement in cure rates and TSR throughout the study. This may be attributed to the URBAN DOTS approach that was implemented in Kampala starting 2012. There was an overall decline in death rate of PBC TB patients that was below the 10% that is reported for all TB cases. This probably meant better chances of survival for PBC TB patients compared to all clases of TB patients combined. Public DTUs reported an overall better improvement in TSR compared to private facilities even though they handled more PBC patients thus highlighting the need to bridge the public private partnership gap in TB management.

The major limitation of our study was that over 26% of all the patients that were documented to be PBCs had unknown/undocumented diagnostic test type. This improper documentation may have misclassified patients to PBC that would have otherwise been classified as clinically diagnosed. The severity of PBC patients diagnosed by GeneXpert was not explored because it was documented by most facilities either Positive or Negative only, therefore, our study could not examine this trend, yet it would have clarified a more representative trend in the severity of PBC TB among patients diagnosed by GeneXpert.

The study sought to analyze and profile PBC TB trends in Kampala over the study period. The proportion of PBC patients increased throughout the years under review. PBC TB patients diagnosed with smear microscopy positivity grading as 3+ consistently contributed to a higher population of those diagnosed as PBC throughout the years compared. This called to attention the consistency in the delayed seeking of care among PBC TB patients and is a stumbling block for effective TB control. We recommend a study of wider scope to estimate how many people a single PBC TB patient is likely to infect with TB per year before s/he is diagnosed and treated. This is crucial in estimating risk and developing strict policies on early detection of PBC TB in this population. GeneXpert testing was a possible contributor to the increase in PBC cases and a challenge to the relevance of microscopy in TB diagnosis [25]. Albeit, we recommend adoption/emphasis of better results reporting options such as MTB detected low, MTB detected intermediate or MTB detected high into the national reporting tools and guidelines used by health workers. This will help capture the severity of the disease at diagnosis in the patient as well as be a marker to late diagnosis/seeking of care.