This study was a prospective, hospital-based and cross-sectional study. The study was carried out at the Paediatrics Emergency Unit (PEU) of FMC, Ido-Ekiti over a period of 6 months (April 2012-September 2012). This PEU is a 13-bedded ward with about 100 patients seen monthly. Federal Medical Centre, Ido-Ekiti is a tertiary hospital that serves as a referral centre for the neighbouring towns in Ekiti State and neighbouring Ondo, Osun and Kwara States. There was an ongoing HIV infection treatment, care and support programme in the hospital at the time of study and at the time of writing.

The subjects were consecutive caregivers of new paediatric patients, aged 6 weeks to 15 years who presented in the PEU with any illness. Only caregivers who gave written informed consents were enrolled for the study. Assent was also obtained from each child aged seven years and above. Caregivers of patients with documented HIV status at presentation were excluded from the study. Each patient was recruited once until the desired sample size was attained.

The sample size was calculated using Fisher’s formula and a prevalence value of 50% giving a minimum sample size of 385 [13]. Five hundred and forty-three caregivers were however counselled for HIV testing.

Caregivers and patients were given HIV pre-test counselling using WHO guideline on PITC with the choice of “opting out” [11]. Counselling cards for paediatric HIV disclosure adopted by Paediatric AIDS Treatment for Africa were used to obtain assent from children aged seven years and above [14]. Pre-test and post-test information were provided in the individual sessions. An interviewer-administered questionnaire designed for the study was used to record information from each caregiver. Data obtained from the caregivers included their age, gender, educational level, occupation, relationship with the child (whether biological parents, grandparents, etc.,) and history of previous HIV testing.

The socioeconomic classification of the children was based on social classification system given by Oyedeji GA [15]. This is based on the mean of a set of scores, assigned for the educational attainments and occupation for the parents. Scores of 1-5 were assigned to each education or occupation category. The scores assigned to the different categories correspond to the 10^th, 25^th, 50^th, 75^th and 90^th percentile of incomes of Nigerians [16]. The mean scores are approximated to the nearest whole number. Scores of 1 and 2 correspond to the upper socioeconomic classes, while a score of 3 is equivalent to the middle class and scores of 4 and 5 to the lower socioeconomic classes.

Ethical consideration: Institutional Ethical Approval was obtained from the Ethics and Research Committee of FMC, Ido-Ekiti. The protocol number assigned to this study is ERC/2011/10/01.

The data were entered into a computer and analysed using the software, Statistical Package for Social Sciences (SPSS) version 15.0 [17]. Categorical variables were expressed in proportions, ratios and percentages. Categorical variables were compared using the Chi-square (χ²) test and Odds ratio where appropriate. Statistical significance was set at p-value less than <0.05.

A total of 543 caregivers participated in the study. [Table/Fig-1] shows the distribution of caregivers in relation to their ages, gender, levels of education, social classes and their relationships with the children. A total of 497 (91.5%) of the caregivers were in the age range of 20-39 years; with the overall mean age of 30.20 (6.86) years. Only two (0.4%) of the caregivers were males. Secondary school education was the highest level of education attained by 337 (62.1%) of the caregivers. Three hundred and fifteen (58.0%) of the 543 caregivers were in social class III. Five hundred and four (92.8%) of the caregivers were the mothers of the children while 12 (2.2%) were accompanied by both parents.

All the 543 caregivers were aware of HIV infection and 538 (99.1%) of them were also aware of where HIV testing was available.

The [Table/Fig-2] shows the association between the ages of the caregivers and acceptance of HIV testing for the children/ward. The association between ages of caregivers and acceptance of HIV testing was not statistically significant (χ²=2.378; df=4; p=0.667). The acceptance rate was 100.0% among the caregivers less than 20 years and those aged above 50 years, while those between 40 and 49 years had acceptance rate of 96.4%, none of these was statistically significant.

The [Table/Fig-2] also shows the association between the levels of education of caregivers and acceptance of HIV testing for the children. There was an overall statistically significant association between levels education of caregivers and acceptance of HIV testing (χ²=18.763; df=3; p≤0.001). All (100.0%) those who had no formal education and only primary education accepted testing, while only 153 (93.3%) of those with tertiary education accepted HIV testing (χ²=16.155; p<0.001). A lower proportion of those with tertiary education accepted HIV testing; this difference was statistically significant (χ²=16.155; df=1; p≤0.001). A total of 153 (93.3%) of the caregivers with tertiary education accepted testing, compared with 377 (99.5%) of those with lower level of education (χ²=16.155; p<0.001).

The relationship between social classes of the caregivers and acceptance of HIV testing for the children are also shown in [Table/Fig-2]. There was an association between level of education of caregivers and acceptance of HIV testing; this was statistically significant (χ²=33.401; df=4; p≤0.001); probably accounted for by much lower acceptance rate among caregivers in the high social class I. For example, only 54 (87.1%) of the caregivers in social class I accepted testing (χ²=28.197; df=1; p≤0.001).

The [Table/Fig-3] shows the association between acceptance rate for HIV testing and the relationship of caregivers with the child. There was an overall association between acceptance of HIV testing and the relationship of the caregivers with the children which was statistically significant (χ²=9.599; df=4; p=0.048). A total of 3 (75.0%) of four uncle/aunties accepted testing. Although the acceptance rate among uncle/aunties was the lowest, this was not statistically significant (χ²=1.761; df=1; p=0.184). Though biological parents and grandmothers had the highest acceptance rate, this was not statistically significant.

The [Table/Fig-4] shows the association between previous screening of caregivers for HIV and acceptance of HIV testing. Previous HIV screening experience of the caregivers was not significantly associated with acceptance of HIV testing (χ²=0.312; df=1; p=0.576).

The association between admission status as an indicator of severity of illness in the child and acceptance of HIV testing is also shown in [Table/Fig-4]. Three hundred and forty-eight (99.4%) of the 350 caregivers of children who were admitted accepted testing, compared with 182 (94.3%) of the 193 caregivers of the children who were not admitted. Thus, the admission status of the children was significantly associated with acceptance of testing (χ²=11.891; df=1; p=0.001).

A total of 291 (95.4%) out of 305 caregivers with prior screening test had their HIV screening done at the Antenatal Clinic, while Voluntary Counselling and Testing, and screening during blood donation accounted for twelve (3.9%) and two (0.7%) respectively.

The overall acceptability rate for HIV testing using PITC Strategy in the present study among caregivers of new patients was 97.6%. This was very high and quite encouraging. The high rate of acceptability was consistent with what had been documented in earlier studies on PITC Strategy in other areas with generalised HIV epidemics, similar to that of Nigeria [3,18]. Wanyenze RK et al., reported acceptability rate of 98% among paediatric in-patient in Uganda [18], Weigel R et al., reported over 95% from Malawi and Mutanga J et al., reported 98.2% from Southern province of Zambia [11,19]. Prior to the study by Wanzenze RK et al., HIV testing was available only upon request and payment [18]. However in their study, as in the present one, HIV testing was free. This might have influenced the high uptake rate. Topp MS et al., demonstrated that introducing PITC using lay healthcare workers in busy urban out-patient departments of primary healthcare centres doubles the HIV testing rate and substantially increased case-finding of individuals with HIV infection [20]. The high acceptability rate supports the view that routine PITC can help in normalising HIV testing and remove a key structural barrier to accessing care and treatment [11,21]. The PITC Strategy is much more likely to reach out to more individuals than the Client-initiated HIV Testing and Counselling that only reaches out to clients who seek HIV counselling and testing services. The use of rapid tests kits for HIV testing also allows prompt delivery of results which enhance timely delivery of antiretroviral drugs with proven efficacy to patients with HIV infection in resource constraint regions.

The high level of awareness of HIV infection and availability of HIV test among caregivers in the present study was consistent with the 2010 report from National HIV sero-prevalence sentinel survey in Nigeria [22]. It was however much higher than the findings of Akpede GO et al., in a study in North-eastern Nigeria on perception of voluntary screening for paediatric HIV and response to post-test counselling by Nigerian parents [23]. The higher awareness rates reported in the National HIV sero-prevalence sentinel survey and in the present study when compared to findings of Akpede GO et al., could be as a result of intense campaign against HIV in the last decade in Nigeria [23]. The high level of awareness of HIV infection amongst caregivers in the present study may therefore be a function of time, which may have also influenced the high level of acceptability of HIV testing recorded.

The low level of previous HIV testing among the caregivers in the present study indicates that there is a need to employ the newer strategies like PITC to get more people to know their HIV status. Factors that influence uptake of HIV testing in the present study, include level of education and social class of the caregivers. Knowledge of these factors are crucial for designing effective control interventions that will curb the spread of HIV infection [24].

The present study also showed that more of the caregivers of children with more severe illness requiring hospital admission accepted testing. Perhaps the severity of the illness made the caregivers to want everything possible done for such children; hence their high level of acceptance.

Although, the findings in the present study suggested that the gender of the caregivers was also an important determinant, with a higher proportion of the male caregivers accepting testing, the interpretation of this finding should be with caution as it may not be a reflection of the reality because of the small number of males involved. It was rather surprising that the findings in the present study suggested that previous HIV screening experience of a caregiver did not influence the likelihood of his/her acceptance for HIV testing. A possible explanation for this was that majority of the caregivers who had HIV testing in the past were offered at the ANC clinic where it was provided as one of the routine investigations which they had to do. The present study, however, offered an opportunity to ‘opt out’, previous studies have shown that this strategy has the advantage of increasing the uptake of HIV testing, increase new case detection and provide clients with the choice to either accept or reject testing in contrast to voluntary counselling and testing [24,25]. The findings in the present study might thus represent the true reflection of acceptance of HIV screening test using the PITC Strategy.

The present study being a hospital based study of may not reflect the acceptability rate for HIV testing in the general community.

Testing for HIV infection using the PITC Strategy in the present study was highly acceptable. The level of awareness of HIV infection was similar to the findings in National sentinel survey. The PITC Strategy should therefore be a feasible strategy for increasing access to paediatric HIV care in health facilities, particularly in generalised epidemic settings like Nigeria. It is therefore recommended that Provider-Initiated Testing and Counselling Strategy should be offered to all children presenting in health care institutions in Nigeria and other African countries.

OAO supervised the study and involved in the data analysis and manuscript writing, OTB studied the conceptualization, data collection and analysis, manuscript writing, LSB involved in the data analysis and manuscript writing. All authors have read and approved the manuscript.