Multiple attempts over the past 20 years to embed IC testing in routine clinical practice have met with significant improvements at individual service level over limited timescales10, but sustaining improvements in testing rates gained through educational interventions and prompts is very difficult. Continuing to prioritise the whole range of ICs is challenging and barriers to testing for clinicians include perceptions of low HIV prevalence and limited knowledge and awareness of the range of ICs11. In Scotland, an IC Testing Action Plan focuses on a smaller number of priority conditions, where system-wide approaches to IC testing can be undertaken.
The conditions included in Stage 1 of the toolkit are:
- Sexually transmitted infections (STI) and blood-borne viruses (BBV)
- Community acquired pneumonia (CAP)
- Constitutional symptoms: unexplained lymphadenopathy, and weight loss and unexplained diarrhoea
NHS Inform has advice (live Feb 2026) for patients that they should be offered an HIV test when they are diagnosed with or investigated for gonorrhoea, syphilis, hepatitis B or C, pneumonia, unexplained swollen lymph glands for more than two weeks or sexually transmitted shigella.
Stage 2 of the IC toolkit will include further detail to support testing in those with, fever, blood test abnormalities (thrombocytopaenia & leukopaenia), infectious mononucleosis and shingles. Discussion with services and laboratories is underway regarding instigating routine testing in these conditions and there are some details that require liaison with the BHIVA/BASHH/BIA HIV Testing Guidelines 2026 writing group – for example the evidence base for including sexually transmitted shigella, salmonella and other enteric infections as indicator conditions. Teaching materials and NHS Inform information will be updated in parallel with implementation of Stage 2 of the toolkit.
What is the evidence for priority conditions?
In Scotland, the Indicator Condition Action Plan (2024), highlights priority conditions for IC testing. These conditions are based on evidence of yield (ie HIV incidence & prevalence), but also applicability, practicality and acceptability of implementation of routine testing. Learning in relation to practicality and applicability continues to develop as the project has progressed – for example there are few or no standard clinical pathways and no public-facing information on unintentional weight loss.
A handful of important recent studies have informed the current revision of the UK HIV Testing Guidelines 3,9,12
- Individuals aged 18–65 years presenting for care to multiple study sites across Europe (including the UK) with one of 14 ICs between January 2012 and June 2014, were routinely offered an HIV test as part of investigation. Data collection on infectious mononucleosis was extended to 2015.
- Of 10,877 patients presenting with an IC and included in the analysis, 303 tested positive (2.8%; 95% CI 2.5-3.1%).
- Analyses included prevalence by region and sensitivity analyses were undertaken.
- The lower 95% confidence interval for HIV prevalence exceeded the accepted threshold of 0.1% in all but 3 conditions.
- HIDES therefore gives us an indication of the likely relative diagnosed prevalence of HIV in different conditions if IC testing were to be effectively applied prospectively in a real-world setting.
- In 6 conditions, the lower 95% CI for HIV prevalence exceeded 2% and these are listed in the table below.
(ref12)
- A retrospective analysis of statutory health insurance data covering 27 million people aged 18-95 years in Germany from 2016-2021.
- A total of 48 ICD codes were examined and grouped into 26 indicator conditions; 15,000 stratified random samples were included for each condition, or all cases where there were less than 15,000 diagnoses.
- Prevalent and incident HIV diagnoses were identified through coding, with stringent criteria to avoid overdiagnosis.
- Diagnoses made in primary and secondary care were included, making the findings likely to be broadly relevant to a UK setting including general practice.
- Analyses allowed for HIV diagnosis prior to, concurrent with and up to 40 months subsequent to the IC and four separate sensitivity analyses were performed.
- The baseline analysis includes both prevalent and incident HIV infections at the time of IC diagnosis (hence including both diagnosed and undiagnosed infections) and Analysis 4, the most parsimonious, includes only those HIV infections diagnosed with HIV within 40 months of the IC but who were without an HIV diagnosis at the time of the IC.
- Population level HIV epidemiology in Germany is broadly comparable to that in the UK, with a very similar total number of people living with HIV. HeLP therefore gives an indication of what the likely diagnosed and undiagnosed prevalence of HIV might be in various conditions.
- It also presents for the first time prevalence data stratified by gender and age ranges which is helpful in considering system-wide interventions to support testing in individual conditions.
(ref3)
A further recent study compares rates of the various ICs in people living with HIV compared to HIV-negative populations:
- A retrospective case-control study of 14,347 people living with HIV compared to 143,470 demographically matched controls using a national communicable diseases database and an insurance registry covering 99% of the population.
- ICs diagnosed within 5 years prior to the date of HIV diagnosis were included.
- All 46 ICs identified were seen at significantly higher rates in people diagnosed with HIV than in matched controls except for mononeuritis, peripheral neuropathy, oral hairy leukoplakia, hepatitis A virus infection, and (in contrast to many other IC studies), hepatitis B virus infection. These findings have informed some changes to the list included in the draft 2026 UK HIV Testing Guidelines.
- Delay in diagnosis from the first identified IC was estimated and was only 15 days (IQR 7-44) in ICs which are AIDS-defining conditions, but 324.5 days (IQR 36-947 days) in those that were associated with immunosuppression but were not ADCs. This latter group includes lymphadenopathy, weight loss and invasive pneumococcal disease.
| Indicator condition |
Strength of Recommendation and Grade of EvidenceA |
Prevalence in prospective studies (HIDES/HIDES 212) |
Prevalence in HeLP retrospective study (Primary analysis)3 |
OR for HIV infection in UK GP-based case control study2 |
Other References |
|
Hepatitis B and C
co-infection
|
1B |
7/72:9.7%
(95%CI 2.9–16.6%)
|
Co-infection not reported |
Small numbers, not reported |
12, 13
|
| Mononucleosis-like illness |
1B <60
2C aged >60**
|
99/2048;4.8%
(95%CI 3.9-5.8%)
|
0.39%
(0.26–0.55)
|
No cases |
9, 12, 13, 14 |
|
Unexplained
lymphadenopathy
|
1B |
31/721: 4.3% (lower 95% CI>2%) |
Not reported |
11.3
(4.5-28.3)
|
9, 12, 15 |
|
Unexplained leukocytopenia/
thrombocytopenia lasting >4 weeksB
|
1B |
16/401: 4.0% (lower 95% CI>2%) |
Not reported |
5.7
(1.4-22.9)
|
9, 12, 13, 16 |
| Pneumonia |
1B
2C for CAP
where not admitted
|
62/1892: 3.3% lower 95% CI>2%) |
0.91%
(0.79–1.03)
|
47.7
(5.6- 404.2)
|
9, 12, 13, 16, 17 |
| Hepatitis C |
1B |
46/1789 (2.6%) |
(Acute) 3.13%
(2.78–3.49)
|
Small numbers, not reported |
2, 9, 13 |
| STI (pooled) |
1B |
31/764 (4.1%)C |
Chlamydia 2.24%
(1.88–2.62)D
Syphilis 11.49%
(10.95–12.05)
Herpes Simplex 0.39% (0.24-0.56)
|
10.8
(2.7- 43.2)
|
16 |
A Evidence Grading from draft (unpublished) BHIVA/BASHH/BIA UK HIV Testing Guidelines 2026
B Confirmed result outside local lab reference range.
C HIDES 1 study16. STI pooled. STI not included as an IC in HIDES2. Significant confounders in that HIV test offer and uptake is much higher in STI clinics than in other settings.
D Males: 7.06% (5.88–8.25), Females: 0.1 % (0.02–0.19)
Clearly the prevalence of HIV in any IC is likely to reflect the background prevalence in the population. Scotland has had great success in recent years in reducing the number of new HIV diagnoses, and in reducing the number of HIV transmissions thought to be transmissions in Scotland to very low levels (37 in 2023 and 24 in 2024). It is likely that the prevalence of HIV in people with IC in Scotland will be lower than that in any published study referred to here. For this reason, we have focussed on conditions in which the lower 95% CI for prevalence was found to be >2% - that is 20x the accepted level for cost effectiveness.
The HIDES 1 and HIDES 2 studies included UK sites (and one of the three largest sites in HIDES 2 was in the UK)12.
HIV transmission elimination
