Socio-demographic, clinical and health facility-related predictors of delayed HIV diagnosis among patients newly diagnosed with HIV in Northwest Ethiopia: a multilevel analysis

Ethiopia recently implemented the ‘test and treat’ strategy for all HIV-infected individuals receiving a diagnosis at the health facility level. However, the impact of this policy in terms of timely HIV diagnosis and factors associated with it were not evaluated. Therefore, this study aimed to determine the magnitude and predictors of delayed HIV diagnosis among newly diagnosed people living with HIV in the northwest, Ethiopia. In this cross-sectional study, a total of 759 newly diagnosed patients were recruited consecutively. The multistage sampling technique was employed to select health facilities and all newly diagnosed patients were included. Delayed HIV diagnosis was dened when there is an established AIDS-dening clinical condition (WHO clinical stage III or IV), irrespective of CD4 count. Data were entered into Epi-Data version 3.5 and exported to STATA version 14 for further analysis. Taking into account the nested structure of the data, multilevel logistic regression analysis has been employed. Four models containing variables of interest were tted. Multivariate multilevel logistic regression analysis was performed to estimate the adjusted odds ratios (AOR) at a 95% condence interval (CI).

In Ethiopia, 729,089 individuals were estimated to be living with HIV by 2018. The annual case of new infections was estimated as 21,606 people and the annual AIDS-related death of 10,960 people in the same year [1,2]. In 2017, only 426,000 were taking Antiretroviral Therapy (ART). Recognizing this low coverage of ART in the country and the prospect of failing to meet the 90-90-90 goal of the United Nations AIDS program (UNAIDS) by 2020, the government endorsed the strategy of universal test and immediate treatment of HIV positive people in Ethiopia [3]. Still, Ethiopia's performance toward the UNAIDS 90-90-90 targets is low [4]. A recent systematic review and meta-analysis on the pooled prevalence of late presentation to HIV/AIDS care in Ethiopia revealed that about 52.89% of HIV patients present late for HIV care [5]. Close to half (45%) of the general population in Ethiopia had never tested for HIV [6].
Early HIV diagnosis and presentation to healthcare facilitates of HIV positive people leads to the advantage of starting treatment as early as possible. Early initiation of ART will decrease disease progression and mortality and enhance HIV prevention [7][8][9]. However, HIV infected people delayed seeking health care and delay in initiating ART in East Africa is quite common [10][11][12]. The situation is similar in Ethiopia in which HIV infected people present to care and treatment within advanced disease stages [13][14][15]. This is worrisome because, patients who present late for HIV care are at increased risk of morbidity and mortality [16,17].
Factors that have been reported to affecting entry into and engagement in the HIV continuum of care include the health status of individuals, exibility of clinic policies, disrespectful treatment from service providers, stock-outs of supplies, stigma, and discrimination, alternate healing systems, diagnosed in hospital, distance to health facilities and poverty [18,19]. Socio-demographic factors associated with late HIV diagnosis include being male, older patients, patients with no formal education and being overweight [20][21][22][23]. Clinical, laboratory and behavioral factors such as fear of stigma, frequent use of alcohol and other substances, presence of symptoms at the time of HIV diagnosis and co-morbidities are also associated with late HIV diagnosis [5,15,19,24,25].
In Ethiopia, though there are several studies about the late presentation of HIV patients for HIV care [14,[26][27][28][29][30], many of the studies may not be generalized because all patients lately presented for HIV care may not necessary, diagnosed their HIV status late. Late presentation to care may occur even though the patients are aware of the HIV status due to several reasons such as fear of stigma [25]. In addition to this, many previously published studies [15,31,32] used a small sample size, were single-site study, used secondary data, and were carried out before test and treat strategy was implemented. So that whether or not test and treat strategy improved HIV test uptake and avoid delayed HIV diagnosis was examined.
Moreover, this study investigated contextual and individual-level factors affecting timely HIV diagnosis of HIV patients over multiple clinical sites. To minimize the confounding effects of nested data over multiple clinical sites, we have employed advanced statistical analysis techniques (multilevel logistic regression) that could improve the limitation of previous studies. Therefore, the study aimed to determine the prevalence and factors associated with delayed HIV diagnosis among newly identi ed people living with HIV in Ethiopia.

Methods And Materials
Study settings and populations The study was conducted in health facilities from three administrative zones of Amhara regional state, northwest Ethiopia. A total of 24 health facilities (9 hospitals and 15 health centers) were selected. Five hospitals and six health centers from East Gojjam Zone, two hospitals and six health centers from West Gojjam Zone, and Bahir Dar city administrations two hospitals and three health centers were selected.
The region has an adult HIV prevalence of 1.2% which was higher than the national average (0.9%) in 2016 [33].
All newly diagnosed HIV positive individuals in the HIV treatment centers of the 24 health facilities who consented to participate in the study were recruited into the study between December 1 st , 2018 and July 30, 2019. The health facilities were included based on an average patient ow of three to ve new cases per month. Participants were eligible if they were 18 years old or over, newly diagnosed with HIV and had no prior exposure to ART. The target population was all newly diagnosed HIV-positive patients in the three administrative zones.

Design and sample size
The study design was cross-sectional. The sample size was calculated using a formula for a single population proportion. The proportion of late presenters to HIV care was taken from a previous study in Ethiopia, which was 52.89% [5]. Using the formula; (Z α/2 ) 2 (P) (p-q)/w 2 ; where Z α/2 =1.96, q=47.11%, p=52.89% and marginal error (w) =0.05. Total=383 And considering a non-response rate of 10% and a design effect of two, a total of 844 participants were required.

Sampling procedures
Three administrative zones were selected purposively. Then a total of 71 health facilities (13 hospitals and 58 health centers) providing ART services were identi ed in the three zones. Health facilities were clustered based on average patient ow as "high load" >3 new patients per month and "low load" <3 new patients per month. A total sample of 844 was distributed for 24 health facilities that ful lled the selection criteria ( g 1). questionnaires. Interviews were conducted in a quiet separate room after obtaining written informed consent. Data collectors were nurses who have received training on HIV treatment guidelines and working in the HIV clinic. All newly diagnosed HIV positive individuals were included consecutively until the required sample size was achieved.

Variables and de nitions
Delayed HIV diagnosis was de ned following the European Late Presenter Consensus working group as a patient who presents for care when there is an established AIDS-de ning clinical condition (stage III/IV) irrespective of CD4 count [36]. Independent variables were categorized into individual and community level factors. Individual-level variables included: age, sex, religion, ethnicity, marital status, educational status, employment status, occupation, wealth status, opportunistic infections, body mass index (BMI), functional status, pregnancy status, any current health complaints. Other individual-level variables were alcohol use, disclosure status, HIV test history, number of sexual partners, condom use, HIV and ART related Knowledge, level of perceived stigma. Health facility-related variables included types of health facility, zones, residence and distance from the health facility.

Data processing and analysis
Data were entered into Epi-Data version 3.5 and exported to STATA version 14 for further analysis. Taking into account the nested structure of the data, multilevel logistic regression analysis has been employed. Four models containing variables of interest were tted using STATA version 14. Model I (Empty model) was tted without independent variables to test random variability in the intercept and to estimate the intra-class correlation coe cient (ICC). Model II examined the effects of individual-level variables, Model III examined the effect of health facility level variables and Model IV examined the effects of both individual and health facility level characteristics simultaneously. The random effects are the measures of variation in delayed HIV diagnosis across communities expressed as ICC and proportional change in variance (PCV). Akaike's Information Criterion (AIC) and Bayesian Information Criterion (BIC) were used to choose a model that best explains the data and the model with low AIC value was taken.
The predicting ability of the model (model accuracy) was evaluated using the Receiver Operating Characteristic (ROC) and area under the curve. These indicated that in model III in which only individuallevel variables were tted it was 81.90% while it was 84.23% in the nal model (IV) that included the health facility level variables. Multivariate multilevel logistic regression analysis was performed to estimate the adjusted odds ratios (AOR) at a 95% con dence interval (CI).

Results
Background characteristics of study participants Prevalence of delayed HIV diagnosis was 25% (95% CI=22-28%). Four hundred thirty-six (57.4%) of the study participants were females. The majority (659, 86.82%) were working functional status while 36.5% of study participants presented with at least one type of health complaint. About 67.6% of them tested for HIV via providers' initiation and 67.3% had never tested for HIV. More than 62% of the study participants had a good level of HIV-related knowledge (Table 1).

Health facility-related characteristics
Three hundred thirty-one (43.6%) were from the East Gojjam zone and 50.2% were from hospitals. The majority (577, 76%) were living in urban areas and 592 (78%) lived in less than one-hour walking distance from the health facility (Table 2).  Table 3).

Results of random effect analysis
In model I (Empty model) the intra-class correlation coe cient (ICC) was 9.26%, in model II (12.30%), in model III (5.45) and model IV (7.70%). The random effects are the measures of variation in delayed HIV diagnosis across communities expressed as ICC and proportional change in variance (PCV). Akaike's Information Criterion (AIC) and Bayesian Information Criterion (BIC) were used to choose a model that best explains the data and the model with low AIC value was taken (Table 4).

Discussion
This study revealed that one-fourth of newly identi ed people living with HIV were diagnosed at an advanced disease stage. Both individual and health facility level factors were found to be associated with delayed HIV diagnosis.
The magnitude of delayed HIV diagnosis was consistent with previous studies from Turkey [21] and China [34]. However, in the current study we have observed that there was an improvement of earlier HIV diagnosis before the advancement of disease compared to reports of previous studies in Ethiopia [5,14,15,26], China [19,24], and South Africa [20,37]. The possible explanations may be time differences between the present and previous studies in which there was a tendency of improved timely HIV diagnosis [12,21,24]. The other reasons may be due to previous studies that de ned delayed HIV diagnosis using CD4 count in addition to WHO clinical stage while this study used only the WHO clinical stage due to the new guidelines [3,38]. This may underestimate the magnitude of delayed HIV diagnosis. Indeed, the WHO recommended using clinical stages 3 and 4 to de ne the advanced disease stage as an alternative to CD4 count [38]. Resuming baseline immunological measurement is important since CD4 count testing still has an important role to play in assessing the baseline risk of disease progression [17,39]. Therefore, in the absence of baseline CD4 level determination, the accurate clinical-stage determination should be improved via training of healthcare providers. Moreover, the limitation of WHO clinical staging to indicate advanced disease stage over CD4 count determination should be acknowledged. A Validation study is recommended to determine the sensitivity and speci city of the WHO clinical stage over immunological measurements.
Baseline clinical pro les of patients such as poor functional status and presented having symptoms at rst HIV diagnosis were associated with delayed HIV diagnosis, which was similar to previous reports [5,15,[29][30][31]. Strengthening existing HIV testing programs including healthcare provider-initiated HIV testing and counseling services should be improved. Furthermore, implementing new programs such as self-HIV testing, house-to-house HIV testing, and resuming HIV testing campaigns would improve timely HIV diagnosis. The reasons for the symptomatic presentation of patients may be attributed to a lack of consistent commitment to implementing community awareness, voluntary counseling, and testing campaigns by the Ethiopian government [4].
This study revealed that a higher level of education was associated with delayed HIV diagnosis, which was similar to the results of other previous studies [29,40]. However, this nding was contrary to other studies that reported an association between a lower level of education and delayed HIV diagnosis [20,21,32,41]. The reasons for the observed discrepancies between delayed HIV diagnoses and the level of education are unknown but may be related to fear of stigma. One study in Ethiopia reported that more educated people are doubtful of HIV test results and repeat it in other health facilities [42]. Such repeat HIV tests may also contribute to delay because repeat HIV-positive testing is associated with a delay in linkage to care. In line with this, patients having a history of HIV test was found to be one of the individual-level factors associated with delayed HIV diagnosis. Interventions targeted at reducing HIV related stigma are highly relevant since a high level of HIV related stigma has an impact on a timely presentation for HIV care [5,25]. The ndings may warrant further investigation of the reason for delayed HIV diagnosis among educated people.
The odds of delayed HIV diagnosis were twice increased among patients diagnosed in hospitals compared to those diagnosed in health centers, which was similar to previous studies [19,37]. The possible explanation may be most patients with AIDS-de ning illness visited hospitals than primary healthcare settings. This might be due to the chains of a referral system from primary to secondary or tertiary units of the healthcare system in Ethiopia. Moreover, previous studies indicated that HIV patients with co-morbid chronic illnesses present late for HIV care [5,28,29]. This may explain why most of the chronic cases are managed at the hospital level. The nding suggested that there should be programs that should be targeting in hospitals and individuals with chronic illness to ensure timely HIV diagnosis and management.

Limitations Of The Study
Though the study was the rst to employ multilevel factors analysis, which contributes to the literature on delayed HIV diagnosis, the ndings should be interpreted within its limitations. In this cross-sectional survey, we were not able to ascertain the causality of associations with delayed HIV diagnosis. Although CD4 counts are important to identify those who presented with advanced HIV disease, we were not able to use CD4 count to de ne delayed HIV diagnosis so that the magnitude of delayed HIV diagnosis may be underestimated.

Conclusion
The magnitude of delayed HIV diagnosis was improved. Here, we recommend using both clinical and laboratory characteristics of a patient particularly, baseline CD4 count and viral load to identify patients diagnosed with advanced disease stage. In addition to individual-level factors, it is important to address health facility-related factors to improve earlier HIV diagnosis. We recommend further large scale studies on the acceptability of new strategies such as self-HIV testing, house-to-house HIV testing and resuming HIV testing campaigns which would improve timely HIV diagnosis. Moreover, public health interventions are important targeted on factors associated with delayed HIV diagnosis. Availability of data and materials

Abbreviations
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declared that there is no competing of interest