*2.2. Setting*

#### 2.2.1. General Setting: Malawi and Lilongwe

Malawi is a land-locked, low-income country in southern Africa with an estimated population of 18 million and with 84% of people living in rural areas [17]. In 2019, the gross national income per capita was USD 380 [18]. Malawi is among the top countries globally with a high burden of TB and HIV/AIDS: in 2019, there were an estimated 27,000 people with TB, of whom 13,000 were HIV positive [19], and there were 1 million people living with HIV (PLHIV) of all ages [20].

Lilongwe is the capital city with a population of about 1 million, according to the 2018 national census [17]. The current study took place in Lilongwe because, at the onset of the epidemic, the majority of cases of COVID-19 came from this city and because partial national lockdown meant it was difficult to travel outside of the city to other regions in the country. Eight health facilities were selected for the study. The selection was made based on high numbers of patients with TB and persons attending for HIV testing and because they were considered by the Lighthouse clinic, the national TB Programme and the national HIV/AIDS Programme to be representative of health facilities within Lilongwe city. The facilities included the central tertiary referral hospital, a secondary referral hospital-specific for HIV/AIDS and including TB patients, one community hospital and five health centres. All these health facilities were in the public sector domain and provided general health services integrated with TB and HIV services. The established staff providing general health and TB and HIV services were used to help with the monthly data collection.

#### 2.2.2. TB and HIV Services

The diagnosis and treatment of TB and HIV/AIDS in Malawi are the responsibility of the National TB Programme and the Department of HIV/AIDS under the Ministry of Health. People with symptoms suggestive of TB (typically these include cough, fever, weight loss, and night sweats) are classified as having presumptive TB when they attend a health facility. Their names are recorded in the presumptive TB register, along with their demographic details. Investigations are carried out according to national and international guidelines [21,22], using sputum smear microscopy and/or the Xpert MTB/RIF (Cepheid, Sunnyvale, CA, USA) assay to establish a bacteriologically confirmed diagnosis of pulmonary TB (PTB). In patients not diagnosed by these methods, clinical assessment, radiography, and other circumstantial evidence are used to establish a diagnosis of clinically diagnosed PTB or extrapulmonary TB (EPTB). Diagnosed TB patients are registered in the TB patient register with demographic and clinical details, given a unique TB registration number and started on anti-TB treatment in accordance with national and international guidelines [21,22]. In brief, patients take their treatment under direct observation by family members or health clinic staff and come to the health facilities once a month to collect drug supplies. The same process is used at all the health facilities. In the selected health facilities in Lilongwe, patients with confirmed or presumed drugsusceptible TB are treated and monitored with the standard 6-month regimen (2 months of

rifampicin/isoniazid/pyrazinamide and ethambutol followed by 4 months of rifampicin and isoniazid), and they were included in this study. Those with drug-resistant disease were not included. Treatment outcomes are monitored, recorded, and reported according to international guidelines [23].

HIV testing is institutionalized in the public health facilities, and provider-initiated counselling and testing are routinely offered to anyone attending for care according to national and international guidelines [24–26]. HIV testing is carried out using rapid testing algorithms. All people diagnosed HIV-positive are referred to ART services for immediate start of ART regardless of their WHO clinical stage or CD4-T lymphocyte cell count.

There is generally good-quality data capture and reporting for TB and HIV/AIDS at all levels due to regular supervision and checking by national programme supervisors. Some selected health facilities also benefit from additional supervision by staff of the Lighthouse clinic.

#### 2.2.3. Data Recording and Reporting for the Study in Health Facilities in Lilongwe

Data were routinely collected on a daily basis by programme staff in each of the eight health facilities in Lilongwe using the standard existing monitoring tools (the presumptive TB register, the sputum laboratory register, and the TB patient register—in which TB treatment outcomes are recorded—and the HIV testing register), most of which were paper based. Moreover, 1–2 weeks after the end of each month, health facility staff collated individual data on TB and HIV variables for the previous month into monthly aggregate data. These were then reviewed and validated by trained data collectors, and the aggregate data were then entered into a data form developed using the EpiCollect5 application (https: //five.epicollect.net, accessed on 4 May 2021). The process each month was supervised by the project country coordinator (HT—appointed for the study).

For TB treatment outcomes, the monthly cohorts of patients enrolled on anti-TB treatment 8 months previously were used—this allowed for 6 months of treatment to be completed and a further 2 months for treatment outcomes to be validated and recorded in the registers. For example, the November 2020 TB treatment outcome data were obtained for the TB patients enrolled and started on treatment in March 2020. National data on COVID-19 cases and deaths reported to WHO on the last day of each month were obtained from WHO situation and epidemiological reports [16]. At the same time as the prospective monthly data were being collected, a schedule and the same procedures were used to collect retrospective data for the previous year. Data were collected on TB and HIV parameters for each month of the COVID-19 period (March 2020 to February 2021) and were also collected for the same parameters for the pre-COVID-19 period (March 2019 to February 2020).

Once all prospective and retrospective data for the reporting month were entered into EpiCollect5, they were checked and validated by the project country coordinator and the overall project monitoring and evaluation officer (PT) based at The Union. Data were then presented in a monthly report as a series of figures and tables to the directors of the national TB programme and national HIV/AIDS programme and to all other relevant stakeholders involved in the project. Any changes made to policy and/or practice at the local health facility or at the national level during that month to counteract the negative effects of COVID-19 were recorded in a narrative table within the report. These monthly reports were always sent and received by the national programme staff within 4 weeks of closure of that month to enable timely surveillance and possible action.

### *2.3. Study Population*

The study population included all patients presenting to the eight health facilities in Lilongwe with presumptive pulmonary TB (PTB), patients diagnosed and registered for anti-TB treatment, and all persons tested for HIV between March 2019 and February 2021: March 2020 to February 2021 was the COVID-19 period and March 2019 to February 2020 was the pre-COVID-19 period. For assessment of treatment outcomes, all TB patients enrolled on TB treatment 8 months previously were considered.

#### *2.4. Data Variables, Sources of Data, and Timing of Data Collection*

Data variables that were collected for TB included aggregate numbers of patients: with presumptive PTB, stratified by male and female, adults ( ≥15 years) and children (<15 years); diagnosed with bacteriologically confirmed PTB by either smear microscopy and/or Xpert MTB/RIF; registered for anti-TB treatment, stratified by bacteriologically confirmed PTB, clinically diagnosed PTB, and EPTB. Standardized TB treatment outcomes of those patients enrolled for treatment 8 months previously were collected and these included treatment success (a combination of those cured with negative sputum smears and those who completed treatment with no sputum smear examination), lost to follow-up, died, failed treatment, or not evaluated [23]. Lost to follow-up is defined as a TB patient who did not start treatment or whose treatment was interrupted for 2 or more consecutive months. Not evaluated is an outcome given to TB patients where no outcome is declared. This includes those who transfer from one facility to another and for whom the final treatment outcome is not recorded. Data variables for HIV included persons who were HIV tested at the health facilities, stratified by male and female, adults ( ≥15 years) and children (<15 years); persons diagnosed HIV-positive; and HIV-positive persons referred to ART. COVID-19 cases and deaths were those reported at the end of each month to WHO and obtained from the WHO epidemiological and situational reports.

Sources of data were the Presumptive TB Register, the Sputum Laboratory Register, the TB Patient Register, and the HIV Testing Register. Prospective and retrospective data for the study were collected between June 2020 and March 2021.

#### *2.5. Analysis and Statistics*

Data were collected in aggregate form, presented as frequencies and proportions, and comparisons were made between the COVID-19 period and the pre-COVID-19 period. The percentage decline in numbers during each month of the COVID-19 period was calculated relative to the numbers during the same month of the pre-COVID-19 period. Comparisons of percentages of persons by age, gender, and test positivity for those presenting with presumptive PTB and being HIV tested as well as the TB registration categories between the two periods were made using the chi-square test and the *p-*values presented. Furthermore, 95% confidence intervals were also presented where appropriate. The relative percentage differences observed between the first 6 months of COVID-19 (March to August 2020) and the second 6 months of COVID-19 (September 2020 to February 2021) were also calculated and presented in the narrative text.
