Page 308 - HIVMED_v21_i1.indb

Page 2 of 8 Original Research

viral infection, accumulation of drug toxins in the body and In addition to the age and gender inclusion criteria above, the
acceleration of age-related degenerative changes by HIV. 4,5,6 cases had to be PLWH and concurrently diagnosed with DM
7
According to the World Health Organization, NCDs may be and/or HTN. All subjects provided written informed consent
present before HIV infection but could be worsened by HIV to participate in the study. Those with a confirmed diagnosis
or the side effects of some of the antiretroviral drugs. other than the NCD-tracer conditions of HTN and DM, as
required for the study, were excluded. Mentally impaired
HIV-NCD comorbidity could have implications for healthcare PLWH and those not on ART were also excluded from the study.
costs, health experiences and survival compared with HIV
alone. Although the effects of HIV-NCD comorbidity on As the ePMS and the Chronic Disease Register are
healthcare costs, health experiences and treatment options maintained separately, the identification and linkage of
have previously been documented, especially in high-income PLWH with the tracer NCDs were time-consuming and
countries, this is not true for most sub-Saharan African laborious. A total of 2969 PLWH and comorbid NCDs
countries, including Zimbabwe. Conducting such studies were identified during a 3-month screening or verification
2
provides valuable information so that targeted intervention exercise among patients registered for the ART programme.
strategies can be developed. Studies on HIV-NCD comorbidity The 3-month period (December 2018 to February 2019) was
are limited to a few cross-sectional surveys that are mainly ideal because all patients were expected to have visited the
based on self-reported data without a comparison group. 2,8 ART sites for their monthly ART supply. After applying the
eligibility criteria, 842 eligible participants from the six sites
HIV-NCD comorbidity causes an additional healthcare were identified. Proportional representation, as guided by
burden. Whilst significant efforts have been made towards the total number of eligible participants per site, was
HIV control in Zimbabwe, untreated or under-treated NCDs followed when selecting participants from rural and urban
can potentially negate the gains achieved through the regions, and from among men and women, within all six
national ART roll-out. Currently, no integrated care for HIV- sites. Simple random sampling was then performed to
NCD comorbidity is available in Zimbabwe. Despite obtain a representative sample per site, coming up with a
accessing free services for HIV and AIDS, PLWH have to pay total of the required sample size.
for the treatment of NCDs, mostly through out-of-pocket
expenditure, which because of the financial burden, may
prevent many patients from seeking care for NCDs. Sample size and sampling criteria
A sample size of 186 (93 per group) was required to detect a
The study sought to determine the effects of HIV-NCD characteristic difference of ± 20% between the two groups
comorbidity, using hypertension (HTN) and diabetes mellitus with a probability of 95% and power of 80%, assuming 50%
(DM) as NCD-tracer conditions, on healthcare costs, health in the control group. A 10% assumption was made for
experiences and treatment options in PLWH in the Gweru missing data and lost to follow-up. As such, a final sample
district of Zimbabwe. size of 208 (104 per group) was needed to ensure sufficient
numbers for analysis.
Materials and methods
Study design Two hundred of those providing informed consent, namely
100 pairs, instead of the initially calculated 208 participants
A repeated-measures, longitudinal quantitative study design or 104 pairs, were enrolled in the study and followed up for
was employed.
6 months. (It is anticipated that the use of 100 pairs instead of
the initially calculated 104 would not significantly affect the
Study setting study power because 100 per group is still larger than the
originally calculated sample size of 93 pairs needed to
The study was conducted at six high-volume government
ART sites that had the highest number of PLWH and achieve the given power at the desired standard error.)
collectively representing over 80% of all PLWH on ART in
Gweru district. The six study sites consisted of four urban The 100 cases were purposively matched by viral load (± 5
and two rural sites. The four urban sites are directly copies per milliliter of blood [copies/ml]), age (± 1 years),
administered by Gweru City Council Health Department, gender (male, female), distance to ART site (± 2 km) and
whilst the two rural sites are administered by the Ministry of geographical location (rural, urban) in the ratio of 1 case : 1
Health and Child Care. control to get a total of 200 study participants.

Study population and participant selection Data collection methods
procedures Quantitative data on participants’ demographic profile,

In this study, HIV patients of either gender, aged ≥18 years, disease-related factors, healthcare costs and health experiences
registered for ART programme in the electronic Patient were collected using an interviewer-administered
Monitoring System (ePMS) and able to respond to the study questionnaire. Health experiences were measured as the
questionnaire in Shona, Ndebele or the English language, number of days spent by participants without carrying out
were considered. usual daily activities because of ill health. The questionnaire

http://www.sajhivmed.org.za 300 Open Access