Quality of data collection in a large HIV observational clinic database in sub-Saharan Africa: implications for clinical research and audit of care

Table 1 Baseline characteristics at ART initiation for patients in the research cohort and routine clinic databases

Characteristic	Research cohort database of patients initiated on ART between April 2004 and April 2005 (n = 559)	Routine clinic database of patients initiated on ART between April 2004 and April 2005 (n = 1233)	Routine clinic database of all patients initiated on ART between January 2000 and December 2007 (n = 6421)	P value ^a
Female, n (%)	386 (64%)	765 (62%)	4112 (64%)	0.417
Age in years, median (IQR)	38 (33, 44)	37 (32, 43)	36 (31, 42)	0.064
WHO stage III & IV, n (%)	496 (89%)	975 (79%)	4581 (71%)	<0.0001
CD4 cells/mm³, median (IQR)	98 (21, 163)	95 (25, 168)	121 (147, 187)	0.232
ART regimen, n (%)
d4T+3TC+nevirapine	414 (74%)	727 (59%)	4224 (66%)	<0.0001
ZDV+3TC+efavirenz	145 (26%)	506 (41%)	2117 (34%)

ART = Antiretroviral therapy, WHO World Health Organization, d4T = stavudine, AZT = Zidovudine, 3TC = lamuvudine, IQR = Interquartile range
a.Based on comparison between 559 patients enrolled into the research cohort and the 1233 initiated on ART over the same time period in the clinic database, but not enrolled into the research cohort. P value derived from Chi-square test across proportions for the categorical variables (female, WHO stage and ART regimen) and Mann-Whitney test for continuous variables (age and CD4+ cell count).