Understanding delays in TB diagnosis: where do people go before accessing TB services?

Four degrees of separation: social contacts and health providers influence the steps to final diagnosis of active tuberculosis patients in urban Uganda.

Sekandi JN, Zalwango S, Martinez L, Handel A, Kakaire R, Nkwata AK, Ezeamama AE, Kiwanuka N, Whalen CC. BMC Infect Dis. 2015 Aug 21;15:361. doi: 10.1186/s12879-015-1084-8.

Background: Delay in tuberculosis (TB) diagnosis adversely affects patients' outcomes and prolongs transmission in the community. The influence of social contacts on steps taken by active pulmonary TB patients to seek a diagnosis has not been well examined.

Methods: A retrospective study design was use to enroll TB patients on treatment for 3 months or less and aged ≥18 years from 3 public clinics in Kampala, Uganda, from March to July 2014. Social network analysis was used to collect information about social contacts and health providers visited by patients to measure the number of steps and time between onset of symptoms and final diagnosis of TB.

Results: Of 294 TB patients, 58% were male and median age was 30 (IQR: 24-38) years. The median number of steps was 4 (IQR: 3, 7) corresponding to 70 (IQR: 28,140) days to diagnosis. New patients had more steps and time to diagnosis compared retreatment patients (5 vs. 3, P < 0.0001; 84 vs. 46 days P < 0.0001). Fifty-eight percent of patients first contacted persons in their social network. The first step to initiate seeking care accounted for 41 % of the patients' time to diagnosis while visits to non-TB providers and TB providers (without a TB diagnosis) accounted for 34 % and 11 % respectively. New TB patients vs. retreatment (HR: 0.66, 95 % CI; 1.11, 1.99), those who first contacted a non-TB health provider vs. contacting social network (HR: 0.72 95 % CI; 0.55, 0.95) and HIV seronegative vs. seropositive patients (HR: 0.70, 95 % CI; 0.53, 0.92) had a significantly lower likelihood of a timely final diagnosis.

Conclusions: There were four degrees of separation between the onset of symptoms in a TB patient and a final diagnosis. Both social and provider networks of patients influenced the diagnostic pathways. Most delays occurred in the first step which represents decisions to seek help, and through interactions with non-TB health providers. TB control programs should strengthen education and active screening in the community and in health care settings to ensure timely diagnosis of TB.

Abstract [1]  Full-text [free] access [2]

Editor’s notes: Delays in tuberculosis diagnosis are well documented in the literature and barriers to diagnosis exist in the community and the health system. This study, in an urban setting in Uganda, included a more in-depth exploration of the steps people take when symptomatic before receiving a TB diagnosis. The median time from onset of symptoms to diagnosis was over two months and the majority of people sought help from relatives and friends and non-TB providers (e.g. private clinics or pharmacies) before accessing public health facilities. There was evidence that having been treated for TB previously reduced the delay, suggesting that existing knowledge of TB in these cases may have influenced their health-seeking behaviour. It was also of interest that delays to diagnosis were greater in HIV-negative people, although it was not possible to establish whether that related to different access to care or to different patterns or severity of symptoms. This is of particular interest because it is thought that in communities with high HIV prevalence, most TB transmission may be attributable to HIV-negative people with active TB.

This study highlights the potential risk of TB transmission in the community during the pre-diagnostic phase. It seems clear that to interrupt TB transmission we need to develop programmes that go beyond public health facilities and this underlines the need for broad community engagement and education as well as more focused programmes to facilitate links between other health care providers and TB services. It is also important to remember that to interrupt transmission we must ensure that TB services and programmes are accessible to HIV-negative people as well as people living with HIV. 

Avoid TB deaths [4]
Comorbidity [5], Epidemiology [6], Health care delivery [7]
Africa [8]
Uganda [9]
  • [10]