Mental Health in Communities Affected by COVID-19 in Uganda

Study design: The investigators will use a multi-stage open-label cluster-randomized trial of
24 randomly chosen villages (n=12 intervention and n=12 control) in Kampala (n=15), Wakiso
(n=3), Mukono (n=4) and Masaka (n=2) districts. Intervention and control villages are at
least 20km apart to avoid contamination

Study procedure:

(i) Randomization and sampling: The investigators shall employ an open label
cluster-randomized trial design, with each village as a cluster, to conduct the study in 24
villages (12 intervention and 12 control villages) in Kampala (n= 15), Wakiso (n= 3), Masaka
(n= 2) and Mukono (n= 4) districts. To ensure balance between the two study arms, villages
underwent stratified randomization on the basis of rural vs urban population. The second
level of stratification will be achieved by a separate randomization procedure performed
within each stratum to ensure gender balance within each cluster

The investigators shall employ the multistage open label sampling approach to select 840
participants from 24 randomly selected villages located in the 362 parishes (5250 villages)
that span the study districts. Study villages will be selected using the probability
proportional to size approach (based on number of parishes/villages per district). Thus, the
investigators shall select 15 villages from Kampala district (3297 villages), 4 villages from
Mukono district (813 villages), 2 villages from Masaka district (436 villages) and 3 villages
from Wakiso district (704 villages). The villages are at least 20km apart to avoid
contamination (individuals interacting during the intervention).

(ii) Participant identification: The investigators shall engage the VHT and village Local
Chairpersons to distribute leaflets describing the study to every 12th household in their
catchment area till they accrue 38 participants in their village who express interest in
participating in the study. This figure is based on an average of 4710 households per
village. Thus 4710/38 is approximately every 12th household. Individuals who express interest
in the study will be given 2 consent forms which they will read and sign should they decide
to participate; they will retain one copy. VHTs will read out the consent form to individuals
who can neither read nor write and obtain informed consent (participants will use their thumb
print soaked in an inkpad to consent). Consent forms will be translated into Luganda. The VHT
will then get the participant's telephone contact and provide it to the project administrator
who will pass them to research assistants. Individuals who don't have phones will be asked to
provide a phone number of a friend or relative through which the interviews can be conducted-
interviews could also be conducted using the VHTs' phone if the participants comfortable
receiving calls on the VHT's telephone. The VHT will inform the project administrator of
participants without personal phones and indicate how that participant can be contacted. The
investigators will conduct a times and motions assessment to document the duration of the
interventions as well as the amount of money spent on transport refunds for the VHTs and the
airtime costs

(iii) Baseline data collection: Trained research assistants will call participants and
schedule a phone interview and collect demographic (age, gender, education level and place of
residence) information and administer the Mental health Knowledge Schedule (MAKS) to assess
for MHL, Kesler-10 (K-10) to measure PD, the Generalized Anxiety Disorder scale (GAD-7) to
assess for GAD, Patient Health Questionnaire-9 (PHQ-9) to assess for MDD Post-Traumatic
Stress Disorder- Primary Care Screen (PTSD-PC) for PTSD and the Alcohol Smoking and Substance
Involvement screening Test (ASSIST) to assess for SUD.

(iv) The intervention: Within a week of the baseline surveys, VHTs in the 12 intervention
villages described above- (a) basic information about the symptoms of PD (b) their adverse
effects, (c) the signs and symptoms of CMD, (d) the benefits associated with having them
treated and (e) the contact details of the research team should they need more information.
See a template of the IEC materials attached. Previous studies have administered the
post-test questionnaires a week after the baseline assessment .

In the 12 control villages, VHTs will distribute the MOH IEC materials that contain
information about (a) the signs and symptoms of the COVID-19 and the preventive measures, (b)
the fact that individuals may suffer from PD during the pandemic (c) the sources from where
they can get psychosocial help and (d) the contact details of the research team should they
need more information. The IEC materials will be distributed to individuals who participated
in the baseline surveys

Mitigating the possibility of contamination between the study arms: Being an educational
intervention, there is a chance that participants in the control villages will access IEC
materials distributed to those in the intervention arm, leading to contamination of the study
findings. To avoid contamination, the investigators will do the following: (i) randomize
villages so that they are at least 20km apart (with the limited human movements during the
COVID-19 pandemic, we anticipate that there will be little by way of people sharing the IEC
materials), (ii) conduct the assessments within a week of distributing the materials to limit
the chances that the information will be widely circulated, (iii) ask participants in the
control arms if they accessed the IEC materials meant for the control group.

(v) Follow-up data collection: Within a week after the VHTs have distributed IEC materials,
research assistants will call participants from both arms and (i) confirm whether they
received IEC materials (and read it), then (ii) schedule a phone interview to administer the
instruments administered at baseline (MAKS, K-10, GAD-7, PHQ-9, PTSD-PC and ASSIST).

Study measures: Research assistants will administer standardized questionnaires to collect:
i) Demographic variables including age, gender, physical address, contact information,
marital and employment status, education level ii) primary study variable (MHL using MAKS)
and (iii) secondary variables including PD using the K-10, GAD using the GAD-7, MDD using the
PHQ-9, PTSD using the PTSD-PC and SUD using the ASSIST.

Participant Follow-up: Participants will be assessed for the outcomes of interest at baseline
and within a week (7 days). Participants who have been referred to the psychosocial teams
will be called after 14 days to find out whether they followed up on the referral and if not,
state the reasons for not doing so.

Adverse event reporting during the interviews: The investigators anticipate that there will
be minimal adverse events that are directly related to the study, mainly psychological
distress resulting from answering the questions during the survey or loss of private data.
However, in the event that the investigators realize any severe form of distress, or loss of
privacy/confidentiality, then the investigators will report it promptly reported to the
SOMREC/UNCST.

Community tracing: In the event that a participant cannot be accessed for the second
interview after the baseline one, then, the investigators will ask the VHT to visit the
physical location and find out the reason for non-response. The VHT will be provided with a
Log to record this kind of information which could be loss of a phone, change in phone number
of withdrawal from the study among other reasons. If the participant can't be contacted on
phone and physically, (may have moved house), then the investigators will consider this a
loss to follow up.

Sample size and statistical considerations Hypothesis: The investigators hypothesize that (a)
the MHL levels are low in the communities in comparison to those observed in high income
countries, (b) a psycho-education intervention delivered by VHTs will lead to a 50% increase
in MHL levels in the intervention arm compared to the control arm (that this difference will
be clinically significant), (c) more participants in the intervention arm will contact the
study team compared to the control arm - and that this difference will be clinically
significant

1. Sample size for objective 1. To document the baseline community MHL level, the
investigators will use the Leslie-Kish formular for calculating the sample size for
cross sectional studies. The formula states that: n = Z2 p (1-p)/E2. n- sample size,
E-The standard error (5%), p- Proportion of patients; 50%.The prevalence was estimated
at 50% as the conservative estimates for outcomes or effects sizes with limited
literature to refer to (in this case, prevalence of poor MHL in the Ugandan community).
Z- The standard normal deviation of 1.96 corresponding to 95% confidence interval
Substituting; n=1.962 X 0.5 X (1-0.5)/0.052 n= 384. b) For the population survey, the
investigators will use the sample size calculation, but use a community correction
effect of 2. Thus the sample size for the community sample will be 384X2 = 764. The
investigators will add a 10% sample size to cater for incomplete data mainly due to
failure to complete an interview. Thus, the total sample size will be 840 participants
(420 in the control and 420 in the intervention arms).

2. Sample size for objective 2: To determine the effectiveness of the psycho-education
intervention, the investigators have calculated the size of effects that the sample will
be able to detect a clinically significant difference in MHL level between the
intervention and control arms. The nature of the educational intervention makes
randomization by clusters (villages) more appropriate in comparison to randomization by
individuals or household.

To estimate the number of villages required per arm for a clinically significant effect size
of 50%: The investigators assume power of 80% at the 95% confidence level, mean MHL score of
14.14 and standard deviation (SD) of 2.19, cluster-size of 35 participants and coefficient of
variation of 0.25, which yields ~8 villages per trial-arm. The investigators shall include a
minimum of 12 villages per arm (50% increment) to cater for the following limitations; i) the
large proportion of participants who do not own mobile phones (~30%) (Incomplete data due to
participants not being able to complete interviews). Therefore, 24 villages (both arms) will
be required; thus, including up to 840 participants (420 per trial-arm).

Data collection and management: Data will be collected on paper and then entered into a
centralized, web-based data management scheme using the open source software REDCap™.

Statistical Analysis

Objective 1: The investigators will report frequencies and percentages of the study outcomes
and their 95% confidence intervals.

Objective 2: An intention-to-treat analysis will be conducted to compare the groups at
baseline and within 4 weeks to assess the effects of the intervention on MHL. The dependent
variable (MHL) will be calculated as a continuous variable. Independent variables including
presence of PD, MDD, GAD, PTSD or SUD will be presented as continuous and categorical
variables. Baseline characteristics of the intervention and control arms will be compared at
the 5% level to assess if successful randomization was achieved. Data on potential
confounders and effect modifiers, including variables that fail to achieve successful
randomization (e.g. socio-demographic parameters) shall be used to control for confounding
and effect modification. Between-subject analysis at week 4 will be used to assess the direct
effect of the intervention by determining if there is a significant difference between the
mean MHL scores in the intervention and control arms. Within-subject analysis will be
performed among patients in the intervention arm by applying the Generalized Estimating
Equations method on repeated measures data.

Data safety management board: The data safety management board (DSMB) will be comprised of
Prof Eugene Kinyanda (Psychiatrist and Senior Epidemiologist), Dr Moses Ocan (pharmacologist
and IRB member), and Dr. Paul Bangirana (Clinical psychologist with vast experience in
conducting RCT) from Makerere University. The PI will have periodic meetings with the DSMB
and provide them with reports about the study activities using an agreed template. There will
be at least two meetings during the study period one at the beginning and another midway
through the study. Members of the DSMB will make a visit at the Butabika Hospital where data
will be collected (phone calls) and stored. During this visit, the study procedures (SOPs and
manuals), study responsibilities (delegation log), site facilities (call rooms/data room),
study materials (stationary and administrative materials) and the recruitment and retention
plan will be discussed.