Table 5.

Results of meta-regression analyses to identify associations and sources of between-study heterogeneity in syphilis prevalence in the Middle East and North Africa (MENA)

Factors		Studies	Samples	Univariable analyses				Multivariable analysis*
		Total N	Total n	OR† (95% CI)	P	P‡ of LR test	Variance explained R2 (%)	AOR† (95% CI)	P	P§ of LR test
Country/subregion:‖
Eastern MENA	Afghanistan, Iran, Pakistan	28	10 865	1.00		<0.001	31.52	1.00		<0.001
Egypt, Jordan, Yemen	Egypt, Jordan, Yemen	4	881	0.89 (0.15-5.10)	0.893			0.66 (0.13-3.28)	0.609
North Africa	Algeria, Morocco, Sudan, Tunisia	48	12 394	5.34 (2.45-11.61)	<0.001			5.01 (2.37-10.61)	<0.001
Horn of Africa	Djibouti, Somalia, South Sudan	26	5629	21.63 (8.89-52.69)	<0.001			6.40 (2.45-16.69)	<0.001
Year of data collection¶		106	29 769	0.88 (0.85-0.91)	<0.001	<0.001	34.61	0.93 (0.88-0.98)	0.005	0.005
Infection type	Current	34	7103	1.00		0.515	0.00	–	–	–
	Ever**	50	9968	1.25 (0.52-3.00)	0.622			–	–	–
	Unclear	22	12 698	0.69 (0.23-2.04)	0.501			–	–	–
Diagnostic method	RPR/VDRL & TPHA/FTA-ABS/RDT	29	6095	1.00		<0.001	22.44	1.00		0.444
	RPR/VDRL	4	488	0.09 (0.01-0.61)	0.013			0.76 (0.15-4.00)	0.746
	TPHA	28	1781	2.17 (0.86-5.45)	0.099			1.29 (0.54-3.07)	0.558
	RDT	23	8707	0.17 (0.06-0.45)	<0.001			0.46 (0.18-1.18)	0.104
	Not specified	22	12 698	0.43 (0.16-1.16)	0.094			0.75 (0.24-2.33)	0.614
STI ascertainment	Biological assay not reported	23	13 066	1.00		0.284	0.15	–	–	–
	Biological assay explicitly indicated	83	16 703	1.66 (0.65-4.20)	0.284			–	–	–
Sample size	<100 participants	42	1960	1.00		<0.001	20.02††	1.00
	≥100 participants	64	27 809	0.16 (0.08-0.32)	<0.001			1.60 (0.62-4.15)	0.329	0.329
Sampling methodology	Non-probability/unclear sampling	66	12 555	1.00		<0.001	18.73‡‡	1.00
	Probability-based sampling	40	17 214	0.16 (0.08-0.34)	<0.001			0.63 (0.25-1.63)	0.339	0.339
Response rate	<60%/unclear	69	18 400	1.00		<0.001	10.23§§	1.00
	≥60%	37	11 369	0.25 (0.12-0.54)	0.001			0.73 (0.29-1.84)	0.495	0.495

AOR – adjusted odds ratio, CI – confidence interval, FTA-ABS – fluorescent treponemal antibody absorption test, LR – likelihood ratio, OR – odds ratio, P – P-value, RDT – rapid diagnostic test, RPR – rapid plasma regain, STI – sexually transmitted infection TPHA – Treponema pallidum haemagglutination assay. VDRL – venereal disease research laboratory

*Adjusted R² in the multivariable model: 48.46%.

†An increment of 0.1 was added to number of events when generating log odds of syphilis infection. This is because 8 stratified measures had zero events.

‡Factors with P ≤ 0.1 were eligible for inclusion in the multivariable analysis.

§Factors with P < 0.05 in the multivariable model were considered as significant predictors.

‖Countries were grouped based on geography and similarity in prevalence levels.

¶Missing values for year of data collection (only one stratified measure) were imputed using data for year of publication adjusted by the median difference between year of publication and median year of data collection for studies with complete information.

**Ever infection indicates seropositivity using antibody testing.

††The high R² was investigated and found to be due to confounding with year of data collection. Most studies with sample size ≥100 were conducted in recent years.

‡‡The high R² was investigated and found to be due to confounding with country and year of data collection. Studies with non-probability sampling were mostly from the Horn of Africa. These studies tended also to be conducted in earlier years.

§§The high R² was investigated and found to be due to confounding with year of data collection. Most studies with response rate ≥60% were conducted in recent years.