Background
Sub-Saharan Africa remains one of the hardest hit regions of the world for HIV infection with about 4.9 % of adults 15 – 49 years living with HIV in 2013 [
1]. Côte d’Ivoire, harbours one of the highest numbers of HIV-infected persons in West Africa with an average HIV prevalence estimated at 3.2 % [2.8–3.8 %] in 2013 [
1]. Vulnerable populations such as female sex workers (FSW) are more likely to be infected with HIV. In sub-Saharan Africa, many countries report an estimated HIV prevalence ≥ 25 % among FSW [
2,
3].
Oral diseases are defined as any morphological or functional abnormalities teeth and supporting structures. These are mainly dental caries resulting from the progressive decalcification of tooth hard tissue, periodontal disease due to inflammation of tooth-supporting tissues and pathologies of the oral mucosa. Oral diseases affect the majority of adult populations worldwide but are particularly high and burdensome in socially marginalized populations due to their limited access to prevention and care [
4]. Several studies in sub-Saharan Africa reported a high prevalence (of the order 40.8 to 86.2 %) of oral diseases among school children and adults [
5,
6]. In Côte d’Ivoire, epidemiologic reports on oral health are scarce, but a high prevalence of oral diseases is expected in vulnerable populations [
7]. In addition to common etiological factors such as poor oral hygiene, consumption of refined carbohydrates and limited access to dental care [
8], HIV infection is a potential determinant of poor oral health. Indeed, several oral-mucosal lesions of viral, bacterial or fungal origin are known to be associated with HIV infection, some of which contribute to the HIV/AIDS clinical staging according to the US Centers for Disease Control (CDC) classification [
9‐
11]. Excessive alcohol consumption, tobacco use and others addictives behaviours associated with a poor oral health are also more frequent among vulnerable populations such as FSW and people living with HIV [
12‐
14]. Therefore, FSW constitute a high-risk population for oral diseases many of which being preventable, treatable or curable. The objectives of this study were to estimate the prevalence of oral diseases including dental caries, periodontal disease and HIV-associated oral-mucosal lesions and to identify their associated factors among FSW in Abidjan.
Methods
Study design
A cross-sectional study was conducted among consenting adult FSW over a three month period, from June to August 2013.
Setting
This study was conducted at the “Clinique de Confiance” in Abidjan, Côte d’Ivoire. The “Clinique de Confiance” is a nongovernmental organization (NGO) dedicated to the overall prevention and care of HIV infection and other sexually transmitted infections among FSW and men having sex with other men (MSM) in the urban area of Abidjan. To date, this NGO ensures the follow-up of more than 1 500 FSW and MSM.
Study population
All adult FSW (≥18 years) who attended this clinic during the study period and who provided their signed inform consent were eligible to participate. About 20 FSW come to the clinic in daily consultation. Since it was not possible to include all the FSW, a number from 1–20 were sequentially assigned to women. Thus, all FSW having an odd number were included in the study in case of acceptance. In case of refusal, the following odd number was selected.
Variables
Oral-health
Dental status was assessed using the Decayed, Missing and Filled Teeth (DMFT) index as described by Klein and Palmer [
15]. The World Health Organization (WHO) defines a standard DMFT scoring scale that qualifies the needs for dental care: very low (0.1 ≤ DMFT index <1.2); low (1.2 ≤ DMFT index ≤ 2.6); moderate (2.7 ≤ DMFT index ≤4.4); high (4.5 ≤ DMFT index ≤6.5) and very high (DMFT index > 6.5). A D index (number of carious teeth) ≥1 indicates at least one carious lesion.
The periodontal status was assessed based on the “Community Periodontal Index” (CPI). The dentition was divided into six sextants: 17 to 14, 13 to 23, 24 to 27, 47 to 44, 43 to 33, 34 to 37. We assigned a numeric code to each sextant based on the examination of 10 indicator teeth (17/16, 1 l, 26/27, 47/46, 31, 36/37). A sextant was considered only if it had at least two functional teeth. The highest code per sextant was reported [
16].
Subjects were then classified into five categories defined by the codes: Code 0 = healthy gum; Code 1 = presence of bleeding; Code 2 = presence of calculus; Code 3 = presence of pockets of 4 to 5 mm; Code 4 = pocket over 6 mm. Presence of pockets was identified by periodontal probing in six sites per tooth (mesio-lingual, centro-lingual, disto-lingual, mesio-buccal, centro-buccal and disto-buccal) [
16]. A CPI score ≥3 was retained to define the presence of periodontitis.
HIV-related oral-mucosal lesions were screened including oral candidiasis, oral hairy leukoplakia and Kaposi’s sarcoma according to Oral HIV/AIDS Research Alliance case definitions [
17].
Oral hygiene level was assessed using the Oral Hygiene Index Simplified (OHIS) [
18]. The OHIS is the sum of the debris index and the calculus index. Both debris index and calculus index are the means of six debris (calculus) scores assessed on six tooth surfaces: 0 = No debris (calculus); 1 = Debris (calculus) on less than 1/3 of the tooth surface; 2 = Debris (calculus) covering between 1/3 and 2/3; 3 = Debris (calculus) covering more than 2/3. The OHIS is assessed by taking six reference teeth (11, 16, 26 of maxillary and 31, 36, 46 in the mandible). The OHIS was categorised into a quantitative ordinal variable with four modalities. Oral hygiene was defined as follows: excellent if OHIS = 0; good if (0.1 < OHIS <0.9); fair if (1 ≤ OHIS ≤1.9) and poor if (OHIS ≥2). In our study, oral hygiene was defined as a three-modality variable by merging the “excellent” and “good” categories.
Socio-demographic and behavioural characteristics
Socio-demographic characteristics including date of birth and education level as well as addictive behaviours (tobacco, alcohol and other drugs consumption) were collected. Tobacco consumption was reported in three modalities “never smokers”, “present or past smokers < 10 cigarettes/day” and “present or past smokers ≥ 10 cigarettes/day”. The level of alcohol use was scored using the Alcohol Use Disorders Identification Test (AUDIT) developed by the World Health Organisation (WHO) [
19]. This 10-item questionnaire assessed alcohol use in the past 12 months based on a 40 points scale. Participants were classified into three modalities: no alcohol user, moderate alcohol user (AUDIT score < 8) or hazardous alcohol users (AUDIT score ≥ 8). Characteristics related to the professional activity of FSW were also collected. These included the number of working days per/week as sex workers, the number of customers per/day in the last working week, use of condom, the existence of a regular partner, the type of sexual practices and the duration of sex work.
The daily number and the average duration of toothbrushing were also collected. The quality of brushing was defined as “low” if the FSW reported brushing her teeth once a day and never in the evening after dinner; “acceptable” if the FSW reported brushing her teeth once a day, in the evening after dinner or at least twice a day before or after meals; “good” if the FSW reported brushing her teeth twice or thrice daily after eating.
HIV infection
The “Clinique de Confiance” systematically offers HIV testing to all FSW attending the clinic, in line with recommendations of the national AIDS program. Dates and results of the last HIV test were registered for all participating FSW. For HIV-infected FSW, additional information was collected including the use of antiretroviral therapy (ART), date of ART initiation, current ART regimen, last known CD4 cell count and CDC clinical stage at first entry into care for their HIV infection.
Data collection
Oral health-related variables were derived from data collected on a standardized form during an oral clinical examination performed by a unique dental surgeon previously trained to the screening of oral diseases, especially oral-mucosal lesions. Oral examination was performed using an overhead dental light. Dental caries were assessed with a probe number 6. The CPI was assessed using a dedicated non-invasive probe with a ball-shaped end of 0.5 mm diameter and a coloured portion of 3.5 mm to 5.5 mm. The dental surgeon was trained for the screening of oral-mucosal lesions using a module designed and developed by the Oral HIV/AIDS Research Alliance [
17]
Variables related to socio-demographic characteristics and addictive behaviours were collected through face-to-face interview conducted by a dental surgeon using a standardised questionnaire before the oral examination.
HIV-related variables were collected by reviewing clinical medical records of HIV-infected FSW participating to the study.
Data analysis
Qualitative variables were described in terms of number and percentage and quantitative variables in terms of mean [standard deviation (SD)] and median [interquartile range (IQR)]. Comparisons were performed using Student’s t test, nonparametric Mann–Whitney U test, or analysis of variance for continuous variables. The Pearson’s Chi-square test or the Fisher’s exact test was used to compare frequencies when appropriate.
Three unconditional logistic models were used to identify factors associated with the presence of dental caries, periodontitis and oral-mucosal lesions. Variables statistically associated with these oral lesions with a p-value <0.20 in univariate analysis were retained and introduced in the initial multivariate model. A backward stepwise regression procedure was then applied to select the final multivariate models. Adjusted Odds Ratio (aOR) were estimated with their 95 % confidence intervals (CI). A p-value < 0.05 was considered for statistical significance.
All analyses were performed with Stata software (StataTM 9.0 College Station, Texas, USA).
Ethical aspects
Before being included, all FSW were informed about risks and benefits associated to their participation in the present study. All FSW provided their written consent to participate. Data collected through this study were part of the IeDEA study Protocol in Côte d’Ivoire (
http://mereva.isped.u-bordeaux2.fr/iedea/) already approved by the National Ethics Committee of Côte d’Ivoire. Oral examination was part of the routine clinical care delivered at the “Clinique de Confiance” during the study period. Regardless to their study participation, all FSW attending the clinic were able to access to this oral examination, if needed (Additional file
1).
Discussion
This study conducted among FSW followed in the urban area of Abidjan, Côte d’Ivoire showed a high prevalence of oral diseases. The prevalence of HIV infection was high and significantly associated with the presence of periodontitis and oral-mucosal lesions. The prevalence of HIV infection was concordant with a previous reports from Côte d’Ivoire where the prevalence of HIV among FSW ranged from 26.6 to 33.0 % [
3,
20]. The high prevalence of oral-mucosal lesions among HIV-infected FSW is largely related to the presence of infections known to be associated with HIV such as oral candidiasis. Despite the lack of precision reflected by the wide confidence interval, the elevated odd ratio showed the particularly high association between the presence of oral-mucosal lesions and HIV infection. Immunodeficiency is a known risk factor for the occurrence of oral-mucosal lesions [
21]. Indeed, HIV-infected FSW with oral-mucosal lesions had a significantly lower number of CD4 cell count compared to HIV-infected FSW without oral-mucosal lesions. Therefore, the prevalence of oral-mucosal lesions in our population of FSW accessing to ART was lower compared to previous reports conducted among HIV-infected persons with limited or no access to ART. In South Africa, Bhayat et al. reported that 53 % of HIV-infected subjects naive to ART presented with oral-mucosal lesions [
22]. In Nigeria, Taïwo et al. reported that before 2005, 36.4 % of HIV-infected women in Nigeria presented with a clinical diagnosis of oral candidiasis [
23]. However, the prevalence of oral candidiasis in our study was higher than the one reported by Owotade et al. in South Africa where 12.8 % of HIV- infected women on ART were diagnosed with oral candidiasis [
24]. The relatively high prevalence of oral-mucosal lesions (23.8 %) in our HIV-infected population accessing ART raised questions concerning ART efficacy or adherence that need to be explored.
The prevalence of carious lesions was elevated in our sample, although not associated with HIV infection. The DMFT index was moderate, with a high prevalence of carious lesions. In addition, few FSW were classified as having a good oral hygiene. The DMFT index reported in our study was higher than the DFMT index reported (2.7) in the study conducted in 2004 by Samba et al. in the general population of Abidjan from 18 to 59 years [
25]. This study included all socio-professional strata among whom junior and senior executives had a more accurate perception of their own health and recognized the need to look after their oral health. Indeed, the number of dental consultations increased with the level of education [
26,
27]. The high prevalence of periodontitis found in HIV-infected FSW was in accordance with a report from N'diaye et al. conducted in Senegal among FSW. In their study, the prevalence of periodontitis was 26.1 % in HIV- infected FSW not on ART and 2.5 % in HIV-negative FSW [
28].
The presence of periodontitis in our study was significantly associated with HIV infection as already reported in industrialized countries [
29,
30]. In 2013, Pattrapornnan P. et al. reported an association between the occurrence of periodontitis and a decline in CD4 counts [
31]. In our study, nearly half of HIV-positive FSW had CD4 count <350 cell/mm
3, which might partly explain the high prevalence of periodontitis. However, mechanisms leading to periodontal disease in HIV-infected people were multiple [
32‐
34]. Another known factor associated with periodontitis is tobacco use [
35]. The absence of association, in our study, between tobacco use and periodontitis might be related to the relatively short median duration of smoking. Indeed, the impact of tobacco smoking on the periodontal tissue is dose-dependent and needs time to induce periodontitis [
36‐
38]. Tobacco and alcohol consumption were particularly high in our study. The prevalence of tobacco smoking reported here was higher than the estimated 1.1 % prevalence reported in the general female population in Côte d’Ivoire in 2002 [
39]. Our results were also higher than those reported by Jaquet et al. among HIV-positive women on ART in west Africa [
13]. This difference might be linked to socio-behavioural factors specific to sex work, which are frequently associated with the use of addictive product [
40]. Therefore, it is important to perform a longitudinal follow-up of this population that might develop these diseases later. In addition, awareness and prevention measures should be directed towards excessive alcohol and tobacco use. Stress was another factor associated with the presence of periodontitis in previous reports. Stigma and discrimination associated with HIV/AIDS are widespread in Africa. FSW live in an extremely difficult climate of stigma related to sex work profession and HIV status as well as human rights violations [
41]. Stress related to stigma might partly explain the association between periodontitis and HIV infection.
Our study has several limitations. The clinical diagnosis of oral diseases, especially oral-mucosal lesions might be operator-dependent. Prior to the study, a training on the diagnosis of oral manifestations of HIV infection was followed by the dental surgeon who performed the clinical oral examination. This training module was designed and developed by the Oral HIV/AIDS Research Alliance (
http://www.aidsresearch.org/). There may be a potential biases in the assessment of oral diseases. In particular, periodontitis was assessed through a specific method developed by the WHO. The existence of alternative methods might limit comparisons with other studies. The study sample might not be representative of FSW in Abidjan. Indeed, FSW already in care for their HIV infection might be more likely to attend the clinic. Nevertheless, there were no particular criteria for FSW to attend the clinic. In addition, the prevalence of HIV infection reported in our randomly selected sample is quite similar to previous findings among FSW in Abidjan and elsewhere in sub-Saharan Africa. The “Clinique de Confiance” is the largest centre for health promotion, prevention and care targeting FSW in Côte d’Ivoire. For logistical reasons, we were also unable to document several important factors such as exposure to refined carbohydrates or HIV viral load in HIV-infected FSW which might have played a major role in the occurrence of oral lesions. Finally, the cross-sectional nature of our study did not allow us to draw any causal relation between oral diseases and their associated factors.
Conclusion
This study showed a high prevalence of oral diseases among FSW in Abidjan. HIV infection was frequent and significantly associated with the presence of periodontal disease and oral-mucosal lesions. Facing the importance of stigma among FSW in sub-Saharan Africa, there is therefore a need to conduct additional studies to assess the impact of stigma on the occurrence of periodontitis in this population.
As oral diseases have a negative impact on the quality of life and can adversely affect various vital functions, there is a need to deliver prevention and care services for oral diseases in most vulnerable population such as FSW in sub-Saharan Africa.
Acknowledgments
We thank the PAC-CI programme team and the West Africa Regional Office of the IeDEA collaboration team for their involvement in the study. We thank the coordinator and all the staff of “Clinique de Confiance” for their open collaboration in the conduct of this study. We give special thanks to all the female sex workers who agreed to participate in this study.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
MN, DGM, PAC, EA, AJ, designed the study. The survey form was developed and validated by MN, DGM, PAC, EA, AJ. Data collection was made and supervised by MN, DGM, PAC, EA, DKE, BKT, CA, SPE, AJ. Analysis and interpretation of data was performed by MN, DGM, PAC, EA, DKE, AJ. The manuscript was drafted by MN, and critical revision for important intellectual content was provided by MN, DGM, PAC, EA,BKT, DKE, FD, AJ. All authors read and commented on the original manuscript and all agreed on the version finalized by MN for submission.
For the IeDEA West Africa collaboration.