Late presentation to HIV/AIDS care at the Douala general hospital, Cameroon: its associated factors, and consequences

Globally, there are approximately 36.7 million people living with HIV, of whom about 70% are in Africa [1]. The burden associated with HIV infection has decreased over the past decade as access to anti-retroviral treatment (ART) has increased. In 2015, WHO recommended that all people living with HIV start ART irrespective of clinical or immune status and Cameroon national guidelines adopted this recommendation in 2016. Despite this change, people living with HIV continue to present to care late and with advanced disease [2]. During the natural course of HIV infection, there is a progressive loss of CD4+ T cells [3–5] and, this leads to a severely immuno-compromised state in the infected host. Unfortunately, in many countries, a substantial number of HIV infected individuals still do not enter health care until late in their infection course [5].

Different criteria have been used to define late presentation (LP). These generally include CD4 cell count and/or AIDS defining disease [6–8]. However a consensus definition has been validated [9]. Late presentation was defined as persons presenting for care with a CD4+ T cell count below 350 cells/mm3 or presenting with an AIDS-defining event, regardless of the CD4 cell count. Advanced Late presentation is presentation for care with a CD4+ T cell count below 200 cells/mm³, or presenting with an AIDS-defining event, regardless of the CD4+ T cell count [9]. LP remains a significant problem worldwide involving both high and low income countries. Reported prevalence in Europe range from 15 and 66% [10, 11] in Asia, 72–83% [5] and in Africa 35–89% [12–16]. Factors associated with LP have been shown to include age, male sex, low level of education, single status, rural residency, fear of stigmas, ignorance and fear of perceived adverse effects of ART [17–20].

LP is associated with significantly heightened levels of HIV related morbidity and mortality [21], increased risk of untoward transmission and high health care costs [22–24]. Surveillance to identify the extent to which LP occurs is crucial and remains inadequate [9]. There is very limited published data in sub-Saharan Africa (SSA), and Cameroon in particular, to the best of our knowledge. In studying the prevalence of LP, we will better comprehend the magnitude of those at increased risk of clinical disease progression. This will also be useful to improve surveillance and assist health professionals in allocation of resources for improved HIV care, especially with the current “test and treat” campaigns in Cameroon.

We therefore aimed to describe late presentation to HIV disease care and to identify its associated factors in a large referral hospital treatment centre between 1996 and 2014.

Of the 3904 files studied, 1710 were incomplete and 194 had less than a year follow up. There were thus 2000 eligible files of which 134 did not have either the initial CD4 cell count and/or the WHO clinical stage (impossibility to define late presentation using our criteria). We finally included 1866 files in our analysis.

Prevalence of late presentation and trend: The overall prevalence of late presentation was 89.7% (1672/1866), 95% CI: 88.2–91.0% and that of Advanced late presentation was 1425/1866 = 76.4% (95% CI: 74.4–78.2%). Figure 1 depicts the prevalence of late presentation and advanced late presentation over calendar time. Late presentation remained very high (above 80%) over the years.

Consequences of late presentation: There were significantly more deaths in late presenters than in early presenters at 3 months (184/1575 = 11.6% vs. 2/189 = 1.1%), 6 months (117/1470 = 7.9% vs. 3/190 = 1.6%) and 12 months (76/1356 = 5.6% vs. 2/187 = 1.1%) respectively (Fig. 2a). The same pattern of death was observed amongst the subgroup that initiated ARV (Table 4 in Appendix). Similarly to deaths, OIs were more frequent in late presenters compared to early presenters at 3 months (1024/1670 = 61.3% vs. 25/192 = 13.02%), 6 months (202/1201 = 16.8% vs. 13/149 = 8.7%), and 12 months (131/991 = 13.2% vs. 14/134 = 10.4%) (Fig. 2b) and the same trends were observed in the subgroup that initiated Cotrimoxazole prophylaxis (Table 4 in Appendix). There was a significant rise (p < 0.0001) in CD4 cell counts in the first 12 months in late presenters. At 12 months median CD4+ cell counts were still lower than in early presenters. In early presenters, there was a slight decrease in the trend of CD4+ T cell counts at 6 and 12 months which was not statistically significant (p = 0.284) (Fig. 2c). The most common OI was pulmonary tuberculosis with respective values of 20.8% (116/1044), 30.1% (64/213), and 25% (36/144) at 3, 6 and 12 months after diagnosis (Table 3).

Infections	3 months		6 months		12 months
	n/N	%	n/N	%	n/N	%
Pulmonary tuberculosis	218/1046	20.8	64/213	30.1	36/144	25.0
Bacterial pneumonia	116/1044	11.1	25/213	11.7	23/143	16.1
Herpes zoster	169/1041	16.2	26/213	12.2	14/143	9.8
Skin Kaposi sarcoma	98/1042	9.4	16/213	7.5	13/143	9.1
Candidiasis	41/1045	3.9	11/213	5.2	13/143	9.1
Cryptococcal Meningo-encephalitis	49/1047	4.7	10/215	4.7	5/144	3.5
Cerebral toxoplasmosis	85/1051	8.1	10/213	4.7	11/144	7.6
Tuberculous meningitis	25/1047	2.4	8/213	3.8	6/137	4.2
HIV encephalitis	27/1047	2.6	1/213	0.50	7/143	4.9
Pulmonary aspergillosis	1/1044	0.10	0/213	0.00	0/143	0.00
Pulmonary Kaposi sarcoma	12/1044	1.2	13/213	6.1	9/143	6.3
Microsporidia	1/1043	0.10	1/213	0.50	1/143	0.70
Isosporidia	3/1043	0.30	2/213	1.0	2/143	1.4
Cryptosporidiosis	2/1043	0.20	0/213	0.00	2/143	1.40

We aimed to describe late presentation, and its associated factors, to HIV care at the DGH between 1996 and 2014. The prevalence of late presentation to HIV care was 89.7% and remained above 80% during the study period. Circumstances of diagnosis (PMTCT program/blood donation, HIV positive partner, routine screening) and employment (student) categories reduced the odds of presenting late, whereas calendar time (2010 above) and diagnosis following clinical suspicion increased the odds of late presentation. Mortality and opportunistic infection (mostly prominently, pulmonary tuberculosis) prevalence remained significantly higher in late presenters at 3, 6 and 12 months.

Following a global consensus, ambitious goals were set for the year 2020, known as the 90–90–90 target [25]. This called for 90% of people living with HIV knowing their HIV status, 90% of people who know their status receiving treatment and 90% of people on HIV treatment having a suppressed viral load [25]. In spite of this, many people still remain unaware of their HIV status and cascade for care for HIV varies widely from country to country and WHO region to region [5].

Knowledge of the prevalence of late presentation can be used as an essential indicator for monitoring prevention programs, and also in assessing the effectiveness of testing campaigns [26]. With set targets, introduction and improvement in ART uptake, the prevalence of late presentation to HIV care did not seem to decrease over time worldwide [2] .

In this study, using the consensus definition for LP, we found a prevalence of almost 90%. This is very similar to findings in a study in neighbouring Nigeria [16]. Our findings were however very high when compared to most other studies in SSA. The criteria for LP were different and thus the importance of using a consensus definition. For example, an Ethiopian study [20] using the criteria of CD4 < 200 found LP to be 61.8%. This is comparable to our own findings if the same definition was used (63.7%). Other authors in SSA found a lower proportion of LP compared to ours as in Malawi 75.4% [27] and Uganda 59.2% [28]. In most of the SSA studies, the sample size was smaller and fewer patients had complete files with data such as CD4 count and clinical stage on entry compared to this study. In European studies, the prevalence of LP using the consensus definition ranged between 50 and 70% [21, 29, 30]. We also found out in this study that the prevalence of LP remained consistently high over the study period. In a Nigerian study, there was a significant decrease in the proportion of individuals presenting late over a 5 year period. However, as in this study LP was 80% and above [16]. This was contrary to a Cambodian study where LP decreased significantly over time from 67% in 2003 to 41% in 2013. These results suggest that in the absence of targeted HIV prevention and diagnostic campaigns, the proportion of HIV patients with LP will continue to be a major problem. Patients will be quite immune suppressed at presentation and as in this study, justifying the high proportion of patients commenced on Cotrimoxazole prophylaxis (77.4%). Cotrimoxazole prophylaxis for toxoplasmosis and pneumocystis in Cameroon is commenced at CD4+ cell counts below 200 cells/mm3.

Several socio-economic factors have previously been shown to be associated with LP [17–20]. Unfortunately we could not evaluate most of them in this study. However, our student population had significantly lower odds of presenting with LP compared to other populations (Employed and unemployed). Similarly those who were employed were less likely to present late than those who were unemployed as found by another author [15]. This finding may be linked to the fact that young people in Cameroon aged 15 to 24 were shown to have better knowledge on HIV/AIDS [25].

Patients who were diagnosed because of clinical suspicion of HIV infection, had higher odds of LP. Having an HIV test due to illness was also found to be an independent risk factor associated with LP to HIV/AIDS care in Harare and Ethiopia [17, 31]. This could mean that individuals attended testing centres primarily when they developed AIDS related conditions. The Public Health message derived from this finding is that LP might be a direct consequence of late diagnosis of HIV and ultimately late linkage to care and treatment [31]. Patients commonly tested during prenatal care, blood donation, diagnosis of a positive partner or routine screening, were less likely to present late. They are therefore more likely linked to care with a higher CD4+ T cell count compared to individuals tested only upon showing symptoms of HIV. The high prevalence of LP provides additional evidence to shift towards routine testing and linkage to care rather than risk based strategies that might not effectively or efficiently engage individuals infected with HIV [32].

We clearly demonstrated that there were increased numbers of OIs (with Tuberculosis as the leading cause) and mortality in late than in early presenters. The management of patients presenting late is challenging. Some of them may die because of AIDS defining illness even before having the possibility to start ART, and individuals with very low CD4+ T cell count have a higher probability of developing complications on starting ART [30]. Unfortunately, a substantial proportion of people who are late presenters are still dying from the disease [33].

Our results should be treated with caution particularly, with regards to reproducibility in different HIV settings. It could be argued that the tertiary referral setting may receive patients with more severe disease and thus a higher prevalence of LP. However other authors have found similar results [16, 34]. The possible drop in CD4+ T cell counts during infections may lead to overestimates of LP when the consensus definition is applied [26]. Also, when based on the consensus definition, LP may be apparently overestimated, nonetheless it can be used as evidence to reinforce current guidelines advocating “test and treat” regardless of CD4+ T cell counts. Other limitations include: some inconsistencies in the data; most especially, misclassification bias and measurement errors (reliability) as we assessed files retrospectively over calendar time; missing data as variables were extracted from clinical files which gave us very little flexibility in our analyses and limited the amount of variables that could be investigated (such as associated factors and also follow up over a longer period of time than 12 months.) Nevertheless, the strength of this study is related to the large number of patients included and the high quality of the data presented in this setting with availability of CD4+ T cell counts and HIV clinical stage in over 1800 patient files on presentation to care.

Nine out of ten HIV/AIDS patients attending the DGH ATC are late presenters to HIV care. This has grave consequences to the individual. Targeted public health interventions to improve early entry into HIV care are urgently needed if Cameroon is to attain the 90–90-90 objective for 2020.