OUP user menu

Missed opportunities for HIV testing—a costly oversight

P. Read, D. Armstrong-James, C.Y. William Tong, J. Fox
DOI: http://dx.doi.org/10.1093/qjmed/hcq236 421-424 First published online: 5 December 2010

Abstract

Objectives: In the UK, one-third of human immunodeficiency virus (HIV)-infected individuals are unaware of their diagnosis, and of those diagnosed a similar proportion have late stage disease. To address this National guidelines have been introduced promoting HIV testing across all medical specialities. We investigated HIV testing patterns in an inner London area with high local HIV prevalence, to identify missed opportunities for HIV testing and its consequences.

Methods: All human immunodeficiency virus (HIV) tests performed in 2008 at Guys and St Thomas’ NHS Trust virology department were reviewed. Tests were stratified for location of request. Case-note review was carried out on all hospital HIV-positive diagnoses outside the genitourinary medicine (GUM) or screening settings to establish the circumstances surrounding the test, and missed opportunities for previous HIV testing.

Results: A total of 40 883 HIV tests were performed in 36 395 individuals. Three hundred and fifty-four (1%) tested positive. Excluding those from GUM or screening settings, 34 (2.8%) of the 1225 inpatients, 17 (0.3%) of the 5303 outpatients and 68 (1.12%) of the 5746 from primary care tested positive. Nineteen (41%) of 46 evaluable hospital diagnoses had presented to local healthcare services within the previous 12 months, 17 (37%) with an HIV indicator condition, but had not been tested. Of the 5303 outpatient tests conducted, 3148 (59%) were performed by either fertility or renal specialist teams. Other specialties conducted relatively few tests. The mean cost of admission for those diagnosed as an inpatient was £36 625 (range £331–223 000). The total cost for the 12 inpatients, who had presented to services in the preceding year but had not been tested was £439 500.

Conclusions: Despite large numbers of HIV tests as screening tests in GUM and antenatal settings, relatively few tests occurred elsewhere with profound costs. Missed opportunities to access this high-prevalence HIV population is concerning and urgent engagement of primary, secondary and tertiary healthcare systems to increase HIV testing and prevent late-stage diagnoses is underway.

Introduction

In the UK, one-third of HIV-infected individuals are unaware of their diagnosis,1 and one-third of new HIV diagnoses are termed ‘late stage’, i.e. CD4 count <200 cells/ml3. The former is associated with onward HIV transmission,2 while late-stage diagnosis is associated with increased morbidity,3 mortality3 and cost of HIV care.4 To address this National guidelines have been introduced promoting HIV testing across all medical specialities.

Missed opportunities for the diagnosis of HIV within a UK hospital setting was first identified in 20055 and National HIV testing guidelines were introduced in 2008.6 These include the use of HIV-indicator conditions to prompt HIV testing and guidance to HIV testing in non-Genitourinary medicine (GUM) settings. It has been suggested that routine HIV testing in a population with an HIV prevalence of 1% is very cost-effective.7

We investigated HIV testing patterns in an inner London area with the UK’s highest HIV prevalence, to identify missed opportunities for HIV testing and its consequences.

Methods

All HIV tests performed on individuals with the age of ≥16 years using the fourth-generation Abbott Architect Combo Ab/Ag HIV test at Guys and St Thomas’ Hospital from January to December 2008 were evaluated. The location and medical specialty requesting the HIV test was recorded. Repeat HIV tests on HIV-infected individuals were excluded. Tests were defined as being performed in screening settings (GUM, family planning, occupational health and antenatal clinics), hospital non-screening settings (other medical outpatients and all inpatients) or general practice. Casenote review of hospital HIV diagnoses in non-screening settings were carried out to verify that the HIV diagnosis was new, and to record the circumstances surrounding the test, medical history and previous visits to the hospital. If an individual was diagnosed HIV positive as part of a hospital inpatient admission the cost of that admission was retrieved from the hospital accounts department. Such costs are based on standard daily tariffs agreed between the hospital and the local Primary Care Trust and vary according to the case complexity and duration of admission if appropriate. Request forms for positive HIV tests originating in general practice were reviewed to establish the indication for HIV testing.

Results

In 2008, 40 883 HIV tests were performed on 36 395 individuals. Of the 36 395 individuals, 354 (1%) tested positive for HIV. The locations of these HIV requests are shown in Figure 1. The HIV detection rate was 34 (2.8%) out of 1225 for inpatients, 17 (0.3) out of 5303 (0.3%) for non-screening outpatients and 68 (1.12%) out of 5746 for primary care. Of a total of 33 633 HIV tests performed in hospital outpatients, 28 330 (84%) were in the screening settings of GUM (n = 18 872), family planning (n = 1221), occupational health (n = 2040) and antenatal clinics (n = 6197). The remaining 5303 tests were in non-screening outpatient settings suggesting only 16% of all outpatient HIV tests were for clinically indicated reasons. The majority of these non-screening medical outpatient tests [3148 (59%) out of 5303] were conducted prior to procedures such as fertility treatment (n = 1895) or renal replacement therapy (n = 1253). Medical specialties such as gastroenterology (n = 83) or neurology (n = 15) performed relatively few HIV tests. The breakdown of the 5303 HIV tests performed by different outpatient specialties is shown in Table 1.

Figure 1.

Location, number and results of HIV tests.

View this table:
Table 1

Medical outpatient speciality requesting HIV tests (n = 5303)a

TeamNo testsNo. posTeamNo testsNo. pos
Fertility18953Dermatology880
ENT273Gastroenterology830
Gynaecology6923Oncology540
Haematology2723Endocrinology230
Respiratory993Ophthalmology220
Renal12531Urology160
Rheumatology301Neurology150
Accident and Emergency2500Cardiology70
Other4770
  • a279 tests were requested by the HIV team and have been excluded as they were known to be confirmatory HIV tests only.

Inpatient notes were available for review for 29 out of 34 cases. Twelve (41%) out of 29 inpatients and 7 (41%) out of 17 outpatients new HIV diagnoses had previously attended Guys and St Thomas’ hospital within the preceding 12 months. All 7 outpatients and 10 out of 12 inpatients had HIV indicator diseases identified by National guidelines, but had not received an HIV test. Figure 2 summarizes the data described above. Thirteen (65%) out of 20 individuals diagnosed with a CD4 <200 cells/ml and 18 (67%) out of 27 those with a CD4 <350 cells/ml had attended services within the previous year.

Figure 2.

Breakdown of HIV tests.

Indicator conditions included viral infections (human papilloma virus-related disease, Herpes zoster, Hepatitis C and Molluscum contagiosum), haematological abnormalities (all seven missed outpatient diagnoses had lymphopaenia/neutropaenia) and bacterial infections such as recurrent pneumonia.

The median CD4 at diagnosis of outpatients was 213 cells/mm3 (range 42–409) and of inpatients was 64 cells/mm3 (range 4–760). Of the 34 inpatient diagnoses, the median age was 43 years, 59% were men, 51% heterosexual, 38% were of Black African ethnicity and 38% were Caucasian. Eighteen out of 34 had an acquired immune deficiency syndrome (AIDS) defining diagnosis. The median length of admission was 15 days (mean 34 days). The total cost of admissions for those who could have been diagnosed when presenting to the hospital in the preceding year was £439 500. The mean cost of admission was £36 625 (range £331–223 000).

Clinical information was available for 40 (59%) out of 68 HIV-positive tests from general practice; 23 out of 40 were conducted for HIV screening and 17 out of 40 for a variety of HIV indicator diseases.

Discussion

This study has two important findings. First, although a large number of HIV tests are conducted, relatively few tests were requested outside traditional testing settings such as GUM, family planning and antenatal clinics. With an estimated local HIV prevalence of 1% in a population of approximately 800 000, there is an urgent need to increase HIV testing based on national guidelines to reduce onward HIV transmission and late-stage diagnosis.

Secondly, previous contact with the hospital was common and of concern was the number who had presented with an HIV indicator condition but remained untested. HIV indicator condition-driven HIV testing would have prevented over one-third of inpatient admissions and 10 (30%) inpatient AIDS diagnoses in this group. The cost of subsequent medical admissions was very high (£36 625) when compared with the average annual cost of HIV treatment in the UK of £8 999 (95% CI £8593 to £9406).8 The personal and economic cost of missed HIV diagnoses is therefore substantial and of clinical governance concern. Since we were unable to evaluate general practice or hospital attendance outside our catchment area and the population is urbanized and highly mobile, it is probable that we have underestimated the proportion of individuals accessing healthcare in the year prior to HIV diagnosis. A prospective study investigating missed opportunities for HIV testing in all new HIV diagnoses at our hospital regardless of disease stage is planned.

In conclusion, we support the urgent implementation of national HIV testing guidelines throughout the hospital setting and urge support from clinicians, commissioners and Primary Care Trusts to facilitate this.

Conflict of interest: None declared.

References

View Abstract