Costs of accessing HIV testing services among rural Malawi communities

Figures & data

Table 1. Descriptive statistics.

Variable	Regression Inclusion	Expected Direction
Gender	Indicator:
	­ Men (reference group) ­ Women	Men are expected to incur higher costs than women to reflect their higher earning potential relative to women
Age (Years)	Indicator:
	­ 16–19 Years; 20–24 Years; 25–39 Years; 40–64 Years; 65+ Years	Financial productivity is expected to increase with age starting from age 20 hence raising the opportunity cost to testing up to age 65
Education	Indicator:
	­ No Formal education (reference group) ­ Incomplete Primary education ­ Some Secondary Education ­ Complete Secondary Education or higher	Education as a proxy for earning potential, implying that the higher the level of education the higher the cost for testing
Number of Children	Continuous: The participant’s number of children	Number of children is positively associated with any child care costs a participant might have incurred while accessing testing hence increasing the total costs incurred
Test Location	Indicator:
	­ Facility-Based Testing (reference group) ­ Community HTC ­ Other Place	Community-based HTC reduces logistic barriers hence lowers the opportunity cost of testing. Other place testing depends on where the person tested for example, if at home testing e.g., self-testing then lower costs than facility-based testing
Amount of Time Taken to Receive Testing	Continuous: Time taken (including travel) in hours to access HIV testing	The more time taken away from work to seek testing, the higher the cost of testing through lost income
Reason for visiting Testing Centre	Indicator:
	­ Had other reasons for visiting a testing centre aside from HIV testing (reference group) ­ Visited a testing centre specifically for an HIV test	Visiting a testing centre for other reasons aside from HIV testing has potential of economies of scope hence reduced total costs
Wealth Index	Indicator:
	­ Households are ranked into wealth quintiles with the poorest as the reference group	Wealth is a proxy for ability to pay; the higher the wealth quintile, the higher the participant’s expenditure to access testing
District of Residence	Indicator:
	­ Blantyre District (Reference Group) ­ Machinga District ­ Mwanza District ­ Neno District	There should not be difference in costs of testing by district

Table 2. Participant characteristics (n = 749)^a.

		Men (n = 237, 32%)		Women (n = 512, 68%)
	N	Percentage	N	Percentage
Age (Years)	16–19	23	9.8%	52	10.2%
	20–24	35	14.8%	135	26.4%
	25–39	96	40.7%	205	40%
	40–64	63	26.7%	102	19.9%
	65+	19	8.1%	18	3.5%
Education	No formal Edu.	19	8.0%	112	21.9%
	Primary Edu.	160	67.5%	331	64.7%
	Some Secondary Edu.	38	16.0%	57	11.1%
	Complete Secondary or Higher Edu.	20	8.4%	12	2.3%
Wealth Index^b,c	Lowest Quintile	64	27.0%	227	44.3%
	2nd Lowest Quintile	40	16.9%	57	11.1%
	Middle Quintile	28	11.8%	69	13.5%
	2nd Highest Quintile	45	19.0%	70	13.7%
	Highest Quintile	60	25.3%	89	17.4%
Test Location	Hospital/Clinic/Health Centre	148	62.5%	295	57.6%
	ANC Clinic	17	7.2%	106	20.7%
	VCT Centre	24	10.1%	31	6.1%
	Community/Mobile HTC	47	19.8%	74	14.5%
	Other Testing Place	1	0.42%	6	1.1%
Number of Children	Mean (min-max)	3	(0–12)	3	(0–13)
Reason for facility visit	HIV Test	168	70.9%	283	55.3%
	HIV Test + Other Services	69	29.1%	229	44.7%
Time Taken	≤1 h	73	30.8%	104	20.3%
	1–3 h	83	35.0%	181	35.4%
	3–6 h	66	27.9%	182	35.6%
	>6 h	15	6.3%	45	8.8%
District	Blantyre	62	26.2%	147	28.7%
	Machinga	70	29.5%	172	33.6%
	Mwanza	30	12.7%	51	10%
	Neno	75	31.7%	142	27.7%

^a3 Participants had incomplete data.

^bWealth index estimated through undertaking principal component analysis of responses to asset ownership and housing environment.

^cAssets selected in the baseline data did not do well in differentiating the poorest from one another.

Table 3. Direct non-medical and indirect costs by gender and cost category.

		Men (US$)		Women (US$)	Total Sample (US$)
Cost Category		Mean (95% CI)	% of Men	Mean (95% CI)	% of Women	Mean 95% CI	% of Total Sample
Direct non-medical costs	Transport	0.25 (0.15–0.36)	6.6%	0.16 (0.11–0.22)	8.7%	0.19 (0.14–0.24)	7.8%
	Consultation	0.03 (0.00–0.05)	0.8%	0.03 (0.01–0.04)	1.6%	0.03 (0.01–0.04)	1.2%
	Food	0.18 (0.14–0.22)	4.7%	0.13 (0.10–0.15)	7.1%	0.14 (0.12–0.17)	5.7%
	Other	0.05 (0.02–0.09)	1.3%	0.02 (0.01–0.04)	1.1%	0.03 (0.02–0.05)	1.2%
Indirect Costs	Child Care	0.06 (0.02–0.11)	1.6%	0.01 (0.00–0.03)	0.6%	0.03 (0.01–0.05)	1.2%
	Lost Income^a	3.24 (2.45–4.03)	85.0%	1.48 (1.31–1.65)	80.9%	2.03 (1.75–2.31)	82.9%
Total direct non-medical and indirect cost		3.81 (2.91–4.50)	100%	1.83 (1.61–2.00)	100%	2.45 (2.11–2.70)	100%

^aLost Income had a median cost of US$1.37; US$2.06 for men and US$0.96 for women.

Table 4. Multivariable analysis of log-transformed Tobit regression model (Dependent Variable: total direct non-medical and indirect costs).

	Determinants (Reference Category)	Coefficient	95% CI	P-value
Gender	(Male)
	Female	−0.323***	(−)0.457–(−)0.189	0.000
Wealth	(Lowest Quintile)
	2nd Lowest Quintile	−0.049	(−)0.239–0.141	0.613
	Middle Quintile	0.169*	(−)0.024–0.362	0.086
	2nd Highest Quintile	0.003	(−)0.176–0.182	0.975
	Highest Quintile	0.175**	0.007–0.343	0.041
Age (Years)	(16–19)
	20–24	0.411***	0.178–0.643	0.001
	25–39	0.640***	0.406–0.873	0.000
	40–64	0.685***	0.395–0.974	0.000
	65+	0.195	(−)0.169–0.56	0.293
Education	No Formal Edu.
	Primary Edu.	0.013	(−)0.151–0.177	0.877
	Incomplete Secondary Edu.	0.253**	0.017–0.489	0.036
	Complete Secondary or Higher	0.530***	0.198–0.863	0.002
Children	No. of Children	0.000	(−)0.033–0.034	0.982
Testing Location	Facility
	Community	−0.396***	(−)0.571–(−)0.220	0.000
	Other	−0.175	(−)0.858–0.508	0.614
Time Taken	Time (Hours)	0.049***	0.023–0.077	0.000
Reason for visiting	HIV Test + Other
	HIV Test	0.079	(−)0.045–0.204	0.211
District	Blantyre
	Machinga	0.059	(−)0.097–0.214	0.460
	Mwanza	0.350***	0.139–0.560	0.001
	Neno	−0.007	−0.164–0.149	0.927
	Constant	0.208	(−)0.113–0.529	0.164
	Observations			746^a

Note: ***p < 0.01, **p < 0.05, *p < 0.1.

^a3 observations had incomplete data.

Table 5. Multivariable analysis of Two-Part Model on total direct non-medical and indirect costs with first part (logit) and second part (Log-transformed OLS).

Determinants (Reference Category)		Two-Part Model
	logit	Log-transformed OLS
Gender	(Male)
	Female	−0.221	−0.517***
Wealth	(Lowest Quintile)
	2nd Lowest Quintile	−0.196	−0.0113
	Middle Quintile	−0.108	0.398***
	2nd Highest Quintile	−0.168	0.0644
	Highest Quintile	0.342	0.161
Age (Years)	(16–19)
	20–24	0.468	0.610***
	25–39	0.777**	0.964***
	40–64	0.674	1.031***
	65+	−0.323	0.736***
Education	(No Formal Edu.)
	Primary Edu.	0.177	−0.0569
	Incomplete Secondary Edu.	0.430	0.248
	Complete Secondary Edu.	0.951	0.628***
Number of Children	No. of Children	0.0604	−0.0164
Testing Location	(Facility)
	Community testing	−0.946***	−0.204
	Other	−0.820	0.0617
Time Taken	Time (Hours)	0.203***	0.0161
		(0.0530)	(0.0197)
Reason for visiting	(HIV Test + Other)
	HIV Test	0.393*	0.0374
District	(Blantyre)
	Machinga	0.253	0.0857
	Mwanza	0.666*	0.434***
	Neno	−0.190	0.0594
	Constant	−0.0902	−0.118
	Observations	746^a	746^a
Pseudo R^2		0.116
Adjusted R^2			0.1579
Log Likelihood		−335.04519	−847.03399

Note: ***p < 0.01, **p < 0.05, *p < 0.1.

^a3 observations had incomplete data.