This was a retrospective review of routine programmatic INH data of under-6 contacts of bacteriologically positive TB patients. Data was collected across 32 private facilities in Lagos Nigeria. The study took place between July-September, 2020.

The study took place in 32 private facilities (26 private for-profit and six private not-for-profits) spread across 13 Local Government Areas (LGAs) in Lagos, Nigeria. Lagos State is divided into 20 LGAs and has a population of 24 million people[16]. The population of Lagos is about 11% of the total Nigerian population. Each LGA in Lagos is supervised by an LGA TB supervisor who oversees that LGA and responsible for TB surveillance. The delivery of healthcare service in Lagos is also organized at three levels namely primary, secondary, and tertiary. In 2003, the Lagos State TB, Buruli Ulcer, and Leprosy Control Program (LSTBLCP) was established as the sub-national authority responsible for TB surveillance activities at state level. The program was further expanded by engaging the private facilities in 2008. The number of private facilities engaged by LSTBLCP increased from 8 facilities to over 150 private health facilities. Lagos facilities are engaged for TB control under four service schemes, i.e., referral of presumptive TB only, provision of directly observed therapy, short-course treatment only, and provision of microscopy.

The cascade of determining the eligibility of a child contact for INH in Nigeria starts by asking smear-positive adult TB patients about the number of children and age of children residing in the same home. Individuals who respond yes to the presence of children less than 6 years were asked by a healthcare worker to bring them to the hospital or in some cases, visited at home for further evaluation by a community health worker. During home visit or presentation to health facility, healthcare workers conduct routine screening for TB using standardized TB screening tool by asking for symptoms of cough of 2 weeks or more, poor weight gain (defined as weight-for-age less than −3 Z-score, or underweight/ weight-for-age less than −2 Z-score, or confirmed weight loss > 5% since the last visit or growth curve flattening), fever (defined as a body temperature of > 37.5 degree Celsius), or other signs of presumptive TB such as body swelling and lymph node enlargement. Once the above symptoms of TB are excluded, adherence counselling is provided and the healthcare worker offers daily INH at 10 mg/kg/day to a maximum of 300 mg/day for 6 mo[2]. Children screened to have any of the presumptive TB symptoms are offered genexpert test/chest x-ray under the National TB, Buruli Ulcer and Leprosy Control Program guidelines. During INH clinic visits, patients’ caregivers/parents are asked for symptoms of TB, investigated if symptomatic, and assessed for adverse events. HCW then documents the child’s weight and dosage of INH per visit in the INH register/ register for the management of under-6 contacts of TB patients. Lost to follow up has occurred when a child misses 2 mo or more of INH refill visits. Such a child is restarted anytime he or she is back in care. However, INH prescription continues from the last refill without any need to restart for children who miss < 2 mo of refill visits, according to the national guideline. At the end of 6 mo, the INH outcomes are also recorded by the healthcare worker following the case definition by NTBLCP.

Completed INH: A child who received a full course of isoniazid (6 mo/180 doses) in 6-9 mo[2].

Died: A individual on INH who is reported to have died of any cause during INH treatment[2].

Lost to Follow Up: If a patient has taken isoniazid for one or more months, then interrupted for 60 d or more[2].

Stopped INH: An individual for whom the INH has been discontinued/stopped by a healthcare worker due to adverse effects or any other reasons[2].

Developed active TB (Failed): If a person develops active TB disease while on INH[2].

Transferred out: An individual who has been transferred to another facility or region to continue treatment[2].

Non-completion: It was defined as the loss to follow-up, death, developed active TB/failed, transfer out, or stopping INH for any other reason[2].

An index case: It was defined as an individual > 15 years old with a positive smear of sputum for Mycobacteria TB.

Childhood contacts: It were defined as children with 6 years of age or less living and sleeping in the same house or group of clustered houses on the same residential site as the index case for at least 1 mo.

A total of 262 children registered/commenced on INH across the 32 private facilities between July 2015 and May 2019 were included in the study. We calculated sample size using the formula n = a² × b/d², where n = sample size, a = Z statistic for a level of confidence, b = prevalence and d = precision or confidence interval. The level of confidence of 95% is conventional at which the value for ‘a’ is 1.96 and ‘d’ is 0.05. A previous study from Brazilian cohort reported INH completion rate of 10%-13% among children of all age-groups[17]. This equates to a ‘b’ value of 0.13. We arrived at an approximate sample size of 200 participants. A total of 262 children under-six were ultimately included in the study.

A four-stage sampling technique (summarized below) was used to select facilities for this study.

Stage 1: Lagos has a total of 20 LGAs out of which purposive sampling technique was used to select seven high TB burden LGAs out of these 20 LGAs. Selected LGAs include Badagry, Ajeromi, Apapa, Ojo, Ifako-Ijaiye, Alimosho, and Oshodi-Isolo.

Stage 2: Facilities providing TB services were first stratified into private and public. Then, private facilities were selected using convenient sampling technique. This was because of less bureaucratic processes in getting access to data from private than public facilities.

Stage 3: Of the 57 facilities across seven selected high TB burden LGAs, thirty-two (32) were proportionately selected. This means that the facilities were selected in proportionate to the number of private facilities in each of the seven LGA.

Stage 4: Eligible child contacts (less than 6 years) initiated on INH were extracted from the facility INH register.

Participants include HIV-negative children less than 6 years who were eligible for INH treatment and placed on INH between July 2015 and May 2019. Eligibility for the study was determined if a child was below age six years, was in contact with a bacteriologically confirmed pulmonary TB (PTB) patient and has lived in the same household for not less than 3 mo before the index TB patient was diagnosed. A bacteriologically confirmed pulmonary TB patient could either be smear positive or has GeneXpert MTB RIF detected. A household contact was defined as a child aged below 6 years of age who lives or has lived within the household of bacteriologically confirmed PTB case. Data collection took place between July and September 2020.

Inclusion criteria: (1) Informed consent by parent/caregiver; (2) Children under 6 years; (3) Children in close contact with or living with a smear-positive adult TB patient; (4) Started on INH at least 6 mo before the study; and (5) Had no symptoms of TB when commencing INH.

Exclusion criteria: (1) The child has received treatment for TB; and (2) Child contacts not living in the same household with the index cases before the diagnosis and child contacts older than 6 years were ineligible for the study and therefore excluded.

INH outcome represents the dependent variable and can either be INH treatment completed or INH non-completed. Independent variables include patient characteristics and facility characteristics. Patient characteristics include age in months, sex (male or female), screened for TB (Yes/No), diagnosis for TB (Yes/No), commenced on INH (Yes/No), date of INH commencement, date, weight and dosage of INH per visit for months 1-6. Facility characteristics include the name of the facility, LGA location, and for-profit status (private not-for-profit vs private for-profit). Descriptive statistics such as frequency and percentage were used to present study variables. Outcome was categorized as INH completed or INH not-completed based on the NTBLCP case definitions. Associations of INH completion with independent variables were assessed using chi-square. The second outcome variable is the proportion of children initiated on INH that dropped off the cascade at months 2, 3, 4, 5 and 6.

Data were extracted from INH register for under-6 children using a standardized excel based data extraction template that mirrors the register. Information such as the age of the patient, sex, INH start date, weight, dosages of INH at months 1-6 as well as health facility-related information such as for-profit status, and LGA were recorded. INH completion was measured using a standard case definition. INH completion was coded as ‘0’ while INH non-completion was coded as 1. For all the Yes/No questions, Yes was assigned code of 1 and No was assigned 0. In this study, children who have taken INH for 6 mo or more were regarded as INH completers, whereas children with documented evidence of < 6 mo of INH intake were regarded as INH non-completers.

Information collected from the INH register was triangulated and compared with that of the INH card to check for correctness and completeness. Where there were discrepancies between the INH card and INH register, the information in the INH card was taken as the correct one and documented in the excel-based data collection template. To minimize bias, the study comprised private for-profit and private not-for-profit facilities across 13 LGAs from both urban and semi-urban populations, as well as large and small-sized facilities.

Data were collected by six trained data clerks using a Microsoft Excel data collection template that was developed for this study. A 3-d training was held on the processes of data collection, entry, validation, and quality assurance. The study tool was piloted in an LGA outside the study location to test the practical knowledge of the trainees and any emerging challenges were addressed. The objectives of the study were concealed from the data clerks. Daily data reviews were held to assure data quality such as missing information and double counting. Collected data were triangulated with another data source, checked by multiple observers, and approved by the supervisor before analysis.

Data were entered into Microsoft Excel and imported into International Business Machine (IBM) Statistical Package for Social Sciences for Windows, Version 21.0. Armonk, NY: IBM Corp for analysis. Data cleaning was done before the analysis. The analysis contains a summary of patient demographic, clinical, and facility characteristics such as age, sex, TB screening status, TB diagnosis status, date of commencing INH, weight, and dose of INH per visit, type of facility, and LGA using frequency and percentage. Bivariate analyses were conducted to determine the relationship between independent and dependent variables (INH completed vs non-completed). None of the explanatory/independent variables was significantly associated with INH completion, i.e., the P value was greater than 0.05 on bivariate analysis for all variables, hence the analysis did not proceed to the multivariate stage.

As the data for this study were collected from routine programmatic data for INH surveillance, no ethical approval was required. Data were de-identified before the analysis for confidentiality and entered into a password-protected Microsoft Excel database. Permission was obtained from the Lagos State TB, Buruli Ulcer and Leprosy Control program. Verbal consent was taken from the parent/caregiver of the children before placing them on INH.

Table 1 shows the characteristics of participants in the study. A little above average (53.8%) of the participants were female, 95.4% were screened for TB and 0.0% was diagnosed of having TB. The participants’ age ranged from 1 to 72 mo with a mean of 36.01 ± 19.67 mo, and 40.2% were between the ages of 1-24 mo.

Table 2 shows the treatment monitoring of the participants. All (100.0%) of the participants were commenced on INH. Across all treatment months, majority weighed 8-15kg i.e., 65.4%, 60.4%, 55.3%, 55.8%, 57.2% and 54.1% in months 1, 2, 3, 4, 5 and 6 respectively. Similarly, and with respect to INH dosage, 150-200 mg dosage was administered to majority i.e., 58.4%, 57.6%, 59%, 60.9%, 61.1% and 63.9% at months 1, 2, 3, 4, 5 and 6 respectively. In the first month, (40.5%) of the participants received less than or equal 100 mg dose of INH, and at the end of the sixth month, 63.9% of the participants received between 150-200 mg dose of INH.

Many (59.2%) of the participants completed INH treatment (Figure 1). Of the cumulative 107 drop-offs while on INH, largest drop-offs were reported at the end of month 2, 52 (49%) followed by 20 (19%), 17 (16%), 11 (10.2%) and 7 (6.5%) at months 3, 4, 5 and 6 respectively (Figure 2).

Open in New Tab   Full Size Figure   Download Figure

Figure 1 Treatment outcome among children on isoniazid Lagos, Nigeria.

Open in New Tab   Full Size Figure   Download Figure

Figure 2 Attrition rate by month among children on 6-mo isoniazid, Lagos, Nigeria.

By using the X²(Chi square) test, the analysis showed that there was no significant relationship between gender and INH completion, [χ² (1, n = 262) =1.3, P > 0.248], age and completion of treatment [χ² (2, n = 262) = 4.1, P > 0.128], type of facility and completion of treatment [χ² (1, n = 262) = 0.03, P > 0.862]. This means that age, gender, and type of facility did not influence the treatment completion of INH recorded among the children (Table 3).

The study abstracted routine programmatic data collected over several years across 32 private clinics in Lagos, Nigeria and therefore suitable for evaluating the completion of programmatic INH among eligible under-6 children in a high TB burden country like Nigeria. The study can be generalized in similar population and setting with high TB burden like urban Lagos, Nigeria. That notwithstanding, routinely collected data are subject to limitations, such as incomplete data, which we minimized by triangulating data from multiple sources, including INH register and INH Care card. Confounders like caregiver characteristics (education, HIV status, attitude), facility characteristics (patient volume, INH availability, waiting time), patient characteristics (distance to facility, HIV status) and health workforce attributes such as knowledge of INH, training and attitude were not assessed due to the retrospective nature of the study and limited information available in the INH register.

Isoniazid (INH) has been proven to be a useful treatment of latent tuberculosis (TB) but its implementation at country level is poor, especially among under-6 children where it is recommended in Nigeria. Even when children are commenced on INH, its completion is rarely investigated.

Completion of INH is necessary to prevent active TB and the development of INH-resistant TB.

This study aimed to determine the completion of isoniazid preventive treatment among under six children in contact with adults with TB and associated factors under routine programmatic use in Lagos, Nigeria.

Retrospective review of INH treatment records of 262 children initiated on a 6-mo INH across 32 private facilities in Lagos, Nigeria.

Only 155 (59.2%) of the 262 participants initiated on INH completed the six-month treatment. Cumulatively, 107 (41.0%) children were lost to follow-up at the end of the sixth month. Of the cumulative 107 loss to follow-up while on INH, largest drop-offs were reported at the end of month 2, 52 (49%) followed by 20 (19%), 17 (16%), 11 (10.2%) and 7 (6.5%) at months 3, 4, 5 and 6 respectively. The analysis showed that there was no significant association between age, gender, type of facility and completion of INH treatment (P > 0.005).

The huge drop-offs in the first 2 mo of INH calls for innovative strategies such as the use of 60-day INH calendar that would facilitate reminder and early engagement of children on INH and their caregivers in care and across the entire period of treatment.

The study added to the body of knowledge on INH completion among eligible children in Nigeria. The findings will alert policy makers on the burden of INH drop-offs and the timing of any interventions that could address retention in care, particularly within the first 2 mo of initiating INH. Future qualitative studies need to unravel actual reasons for these huge loses while on INH.