Assessment of standard HIV testing services delivery to injured persons seeking emergency care in Nairobi, Kenya: A prospective observational study

Emergency departments (EDs) in Africa are contact points for key groups for HIV testing services (HTS) but understanding of ED-testing delivery is limited which may impeded program impacts. This study evaluated the offering and uptake of standard HTS among injured persons seeking ED care at Kenyatta National Hospital (KNH) in Nairobi, Kenya. An ED population of adult injured persons was prospectively enrolled (1 March—25 May 2021) and followed through ED disposition. Participants requiring admission were followed through hospital discharge and willing participants were contacted at 28-days for follow up. Data on population characteristics and HTS were collected by personnel distinct from clinicians responsible for standard HTS. Descriptive analyses were performed and prevalence values with 95% confidence intervals (CI) were calculated for HIV parameters. The study enrolled 646 participants. The median age was 29 years with the majority male (87.8%). Most ED patients were discharged (58.9%). A prior HIV diagnosis was reported by 2.3% of participants and 52.7% reported their last testing as >6 months prior. Standard ED-HTS were offered to 49 or 8.6% of participants (95% CI: 5.8–9.9%), among which 89.8% accepted. For ED-tested participants 11.4% were newly diagnosed with HIV (95% CI: 5.0–24.0%). Among 243 participants admitted, testing was offered to 6.2% (95% CI: 3.9–9.9%), with 93.8% accepting. For admitted participants tested 13.3% (95% CI: 4.0–35.1%) were newly diagnosed (all distinct from ED cases). At 28-day follow up an additional 22 participants reported completing testing since ED visitation, with three newly diagnosed. During the full follow-up period the HIV prevalence in the population tested was 10.3% (95% CI: 5.3–19.0%); all being previously undiagnosed. Offering of standard HTS was infrequent, however, when offered, uptake and identification of new HIV diagnoses were high. These data suggest that opportunities exist to improve ED-HTS which could enhance identification of undiagnosed HIV.

125 Baseline data included information on sociodemographic aspects, past medical history, 126 characteristics of the injury event, previous health behaviors pertaining to HIV testing and 127 prevention, and use of alcohol and other recreational substances. Participants were followed until 128 ED disposition (completion of ED care) at which time they had information collection on standard 129 HIV testing engagement (offering by clinical staff and uptake by the participant). Data on injury 130 burdens and psychometric data on participant perspectives on ED-based HIV services were also 131 collected at time of ED disposition. Injury characteristics were classified via the standardized 132 Abbreviated Injury Scale, as has been used previously in similar settings (33, 34). The 133 psychometric data utilized a battery of Likert items which were piloted for understandability and 134 reproducibility in the study setting, and assessed with five-point response scales. The items focused 135 on appropriateness of ED-HIV testing and specific aspects of physical space, interaction times, 136 ability to ensure confidentiality and adequacy of patient-provider relationships (Supplement 1).
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 5, 2022. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.  . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 5, 2022. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 5, 2022.

208
The most common mechanisms of injury were road traffic accidents (51.9%) and falls . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 5, 2022.

230
In comparative analyses between those offered and not offered ED-HTS no significant 231 differences in characteristics were identified (Table 3). In evaluation of factors associated with . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 5, 2022. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 5, 2022.   Africa (17, 23). The results demonstrate that there exists substantial potential to 257 increase HTS for those seeking emergency care of injuries, and that the population had frequent 258 uptake of testing when it was offered in the study setting. Given these findings, in conjunction 259 with the relatively high frequency of newly identified PLHIV among those completing testing, 260 development of improved ED-based HIV-programming may represent an impactful approach to 261 enhance testing and identification of undiagnosed individuals from high-incidence populations.

262
In the current data standard ED-based PITC for HIV by existing clinical providers was 263 infrequently offered to those seeking care. This is congruent with the limited data that exists from 264 other emergency care settings in sub-Saharan Africa (17, 18). In a 2018 cohort from South Africa, 265 in which the majority of patients were men and seeking emergency injury treatments, only 25% of 266 patients were offered testing (18). Given that Kenya national guidelines call for universal opt-out 267 PITC at every healthcare encounter (25), the low occurrence of test offering in the studied 268 population indicates that there is potential to increase testing delivery from the ED setting.
269 Furthermore, as emergency care interactions in LMICs are common, with a median of 30,000 visits 270 per facility per year (9), opportunities may exist to augment HTS to large numbers of persons from 271 LMIC ED venues. As injuries are one of the most frequent reasons for receiving emergency care, 272 and those most likely to be injured are also key target populations for testing initiatives the ED 273 venue may be an even more important focus point to improve HIV service delivery (6, 11-14). In 274 Kenya specifically, younger people, who are more frequently injured and also comprise more than . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

292
The proportion of newly identified PLHIV among participants completing testing was 293 more than two-fold greater than the baseline population prevalence in the study setting (24). These 294 data comport with prior studies in which HIV burdens among persons with injuries in Africa have 295 been found to be greater than the prevalence in the corresponding populations (16), and that ED-296 based testing frequently identifies incident infections (18, 19, 37). As the current data pertaining 297 to testing outcomes were limited, secondary to the low numbers of test offered, precise estimates . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 5, 2022. ; https://doi.org/10.1101/2022.05.03.22274632 doi: medRxiv preprint 298 on HIV burdens in the population studied are difficult to make and larger sampling frames are 299 required. However, the high observed disease burden does support the potential to increase 300 identification of PLHIV through engagement and testing in ED settings. Although gains in 301 combating the HIV epidemic require care coordination and treatment following diagnosis (38), 302 identification of PLHIV is a requisite first step, which could be advanced towards with 303 strengthened ED-HTS in sub-Saharan Africa, an approach which has been recognized as important 304 in alternative higher-resourced settings (39).

305
The is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 5, 2022. ; https://doi.org/10.1101/2022.05.03.22274632 doi: medRxiv preprint 321 altered mental status were excluded. Although this could affect the generalizability, the internal 322 validity of the results to the population of interest would not be impacted. As well, the data may 323 not be completely generalizable to alternative settings as they are derived from a clinical site that 324 is relatively well-resourced for the provision of HIV testing and further study is needed to 325 characterize HIV services in settings with differential resources. Related, although an a priori 326 secondary objective was to assess factors associated with the standard offering of HIV testing in 327 the population the low event rate for test offering precluded the ability to complete those analyses.
328 As such, future research aiming to assess opportunities and challenges associated with ED-based 329 HIV testing in high-burden settings, such as Kenya, will be beneficial to inform programming. 338 This prospective longitudinal data found that the offering and delivery of ED-based HTS to the 339 population of adult injured persons seeking care was low, but when offered, there was frequent 340 uptake and a high relative identification of newly diagnosed PLHIV as compared the prevalence 341 in the general population. These findings suggest that there are opportunities to increase ED-HTS 342 delivery, and that if achieved, this could enhance identification of undiagnosed persons during 343 emergency care encounters. Given the potential margin for gains in HIV testing and diagnosis, . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 5, 2022. ; https://doi.org/10.1101/2022.05.03.22274632 doi: medRxiv preprint 344 coupled to the limited available research from LMIC emergency care settings, future studies to 345 better understand of opportunities and challenges in ED-based HIV programming are needed to 346 inform development of more effective and context appropriate testing strategies.

347
348 Acknowledgments: The authors thank all of the personnel who assisted in the research and study 349 participants who took part in the work. As well, acknowledgment to Ananya Suram who 350 contributed to production of data graphics.
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 5, 2022. ; https://doi.org/10.1101/2022.05.03.22274632 doi: medRxiv preprint