More people with HIV in England have dropped out of care than remain undiagnosed

Dr Ann Sullivan presenting at BHIVA 2023. Photo by Roger Pebody.

As many as 20,000 of the approximately 96,000 people in England who are living with HIV may not be virally suppressed, so could potentially be able to transmit the infection to others, the British HIV Association Conference heard in Gateshead last month.

This is an upper estimate. But several presentations at the conference continued a theme introduced at last year’s conference: that the figures for the proportion of people in England who are virally suppressed are over-optimistic because they do not take full account of people who have dropped out of care after initially receiving suppressive treatment.

This ongoing problem gave impetus to suggestions that the anonymity offered to patients who attend HIV clinics could be relaxed, to the extent of attaching their NHS number to their records. This would allow matching patients to GP records, which may be more up to date. It was also pointed out, however, that some people may have, or feel they have, good reasons to conceal their HIV status, particularly in a climate of hostility to immigrants. Monitoring which would alert clinics to patients who appear to have dropped out of their care, combined with funding to help them re-link patients to care, might be a better way forward for now.

This problem is not unique to England. As the proportion of people with HIV who have never been diagnosed or started treatment decreases, the number who drop out of care and who are at renewed risk of becoming ill and passing HIV on becomes more significant. Not counting them also brings into question the reality of progress towards the UNAIDS Fast-Track Targets for 2030.

How many are out of care – and not virally suppressed?

Dr Ann Sullivan of the UK Health Security Agency told the conference that there were several reasons why the UK’s apparently excellent record in diagnosing and treating people with HIV might not be quite as good as it looks.

In terms of the figure reported to global health agencies, the latest one for 2021 is that 95.4% of people with HIV in England know their status, of those 98.7% are on antiretroviral therapy (ART) and of them 97.8% are virally suppressed – meaning that 92% of all people with HIV have an undetectable viral load and cannot transmit HIV. This exceeds the 95-95-95 Fast-Track target at each of its three steps and as a whole: 95-95-95 implies that 85.7% of people with HIV in a country should be in care and virally suppressed by 2030, so England appears to have hit this nine years early.

The problem with this is that when the 90-90-90 and 95-95-95 targets were devised, they assumed that once people are linked to care and on ART they stay that way.

This was a pragmatic decision: establishing if, and when, someone has dropped out of care is far more difficult than establishing if they have started care. That’s because it involves, firstly, deciding how long someone has to miss an appointment for in order to be defined as out of care; secondly, to decide how soon being out of care leads to loss of viral suppression; and lastly and most importantly, finding out whether they are truly out of care or are getting care somewhere else – either in a different clinic or country. The same problem applies to PrEP, which is why most PrEP statistics measure PrEP initiations and not continuation.

Of the 95-95-95 targets, the figures for diagnoses are probably the most reliable as they estimate diagnoses by taking the ratio of positive results to all tests. This latter figure dipped during COVID but generally does not change much. But if the true number of undiagnosed sits at the upper end of the 95% confidence interval (i.e. is the highest likely ‘true’ figure given inevitable statistical fuzziness), the actual proportion of people with HIV diagnosed could be 93.7% rather than 95.4%. That would be 6044 undiagnosed people instead of 4410 – far fewer than the number who may have dropped out of care.

However, if you add on to this the assumption that “missing data equals a negative outcome” then the proportion who are virally suppressed becomes much lower. If all the people recorded as having a positive test result but not recorded as starting treatment are indeed not taking ART then the proportion of the diagnosed people who are on ART drops from 98.7% to 94.1%.

But the figures really fall off if it is assumed that people recorded as starting treatment but not continuing treatment have all stopped treatment and are therefore not virally suppressed. This would reduce the proportion of people diagnosed who are virally suppressed to 82.5% – or to 77.4% if the proportion of people undiagnosed is the higher figure.

Adding in a couple more steps and assuming that all along the way, missing data means that people have dropped out of care, we arrive at the following figures: out of an estimated 95,930 people with HIV in England:

  • 91,520 are diagnosed
  • 89,820 have been seen at least once at an HIV clinic
  • 86,140 have started treatment
  • But only 75,520 are continuing on treatment and therefore virally suppressed (78.7%).

Things are likely to be better than this – but could even be worse. One crucial judgement is the definition of when someone is ‘out of care’. Currently, the definition of out of care is a patient who has not returned for an appointment in 15 months. A patient who last attended 14 months ago would be considered out of care, but once their gap in attendance goes beyond 15 months, they stop being counted and disappear from the data.

Over the last three years we have data for (2019-2021), the estimated number undiagnosed has shrunk, from 5600 to 4400. The number not returning for an appointment within 15 months bulged under COVID, from 3094 in 2019 to 4740 in 2020, but by 2021 had started to shrink again, to 4118. The 'COVID bulge' may have been exaggerated: NHS England’s HIV and AIDS Recording System (HARS) imposes a fairly rigid model – an 'attendance' means physical attendance and blood tests being taken. During COVID, with many patients having virtual appointments and antiretrovirals being couriered to them without tests, some could in fact have been in care.

Using the 15-month definition of 'out of care' reduces the number of people with transmissible levels of HIV to 11,659, or 12,177 if the absence of recorded viral suppression implies people are not suppressed, rather than the positive recording of it. This is still just within the target, with 87.3% suppressed.

What if, instead, you take as your definition of out of care people who haven’t returned for an appointment since 2016, i.e. in the last five years?


virological suppression

Halting of the function or replication of a virus. In HIV, optimal viral suppression is measured as the reduction of viral load (HIV RNA) to undetectable levels and is the goal of antiretroviral therapy.

90-90-90 target

A target set by the Joint United Nations Programme on HIV/AIDS (UNAIDS) for 90% of people with HIV to be diagnosed, 90% of diagnosed people to be taking treatment, and 90% of people on treatment to have an undetectable viral load. 


A patient’s agreement to take a test or a treatment. In medical ethics, an adult who has mental capacity always has the right to refuse. 


The presence of one or more additional health conditions at the same time as a primary condition (such as HIV).


In a case-control study, a process to make the cases and the controls comparable with respect to extraneous factors. For example, each case is matched individually with a control subject on variables such as age, sex and HIV status. 

You might at first think that the number falling out of care would be smaller because some of those not returning for 15 months now have nearly four more years in which to reappear. In fact, however, the number defined as out of care grows dramatically. If you take all the individual patient records in England where someone has gone missing for those five years or more, then the number out of care grows from 4118 to 15,963, implying that the total number of people with HIV in England and not virally suppressed is 28,130 – 29% of all people with HIV, and missing the  90-90-90 target of 72.9% unsuppressed, let alone the 95-95-95 one.

The reason this figure is so much greater is largely because people who drop out of care for longer periods do not continue to be followed up. This could imply – as the 28,130 figure does – that there are a huge number of people around not in care. However, delving deeper into the records finds that a large proportion of those 'not retained in care' – possibly a majority – are in care. Some people really have disengaged from HIV care or have died, but more may have simply moved either their address or just their clinic, without telling anyone. 

Asked to estimate the true number of patients who were in care but have dropped out since 2015, Dr Sullivan said 18,000 would be the upper limit, but that the true number was probably about 15,000 – over three times the number who have never been diagnosed, and over one in six people with HIV.

Tracing and re-engaging people out of care

How do we trace these people and encourage them to come back into care? Dr Kate Childs of King’s College Hospital in south London presented an update of their drive to re-connect missing patients to care. This required charitable funding: Dr Childs emphasised that re-engaging patients was very different work to retaining patients in care. “The mantra of ‘prevent, test, treat, retain’ for managing HIV does not really cover ‘re-engage’ because while ‘retain’ is mainly about support, ‘re-engage’ involves a considerable amount of admin work that probably requires a dedicated post,” she said.

Dr Kate Childs talks about the study she presented at BHIVA 2022 about people dropping out of care.

As an illustration of the work that needs to be put in: the programme initially identified 2275 out of 7092 patients who were out of care across three clinics – just under one-third. The UK Health Security Agency traced 521 who were in care elsewhere, but it took dedicated work by King’s to identify 930 more people who had either died, moved aboard or were in care elsewhere. This left 824 (about one in nine patients) who had potentially dropped out of care in the UK.

Of these, only 153 (under one in five) were successfully re-engaged in care. Of them, a quarter dropped out of care again.

Rose Mower, a Clinical Nurse Specialist from the neighbouring Guy’s and St Thomas’ NHS Trust described the patients her re-engagement project had managed to get back into care. In her NHS trust, 41% of people hospitalised for HIV-related conditions were people who had dropped out of care, compared with just 11% who were newly diagnosed. (At King’s, 63% of those hospitalised were people who had dropped out of care.)

They managed to make contact with 82 patients. Of them, 70% were men (unlike at King’s, where the majority were women); 55% were of Black ethnicity (African or other); 42% had a CD4 count below 350; and 40% had viral loads over 200.

Of the 82 patients, however, only 52% actually re-attended their clinic and of those 82% were virally suppressed. But eight had had to be hospitalised. The majority (72%) had significant co-morbidities, including one-third with mental health issues.

Poverty (often involving an inability to get to the clinic), food insecurity, drug and alcohol problems, poor mental health and childcare issues were identified by both Rose Mower and Kate Childs as being typical issues for patients who fell out of care. The most useful tool for clinicians trying to re-engage these patients, both emphasised, was having a dedicated mobile phone which allowed patients to speak or text directly to the doctor or nurse in charge of re-engagement.

In a separate session on tracing patients who are out of care, there was an even split in the audience about whether linking HIV patients to their NHS number was a good idea. (At the moment, HIV clinics are allowed to offer patients exceptional confidentiality, without linking their data to national patient records, under legislation originally intended for sexual health clinics.) It might scare away patients who cannot be reassured that their details would never be passed on to other arms of the government such as the immigration authorities.

Not assuming that a patient’s lack of consent to contacting their GPs at the time they are diagnosed is permanent, and instead re-checking it periodically, might be a more acceptable way forward right now than linking them to their NHS number. There was also a general acceptance that it would be ethical to state to patients that in cases of life-threatening illness, GPs might be informed. As well as directly trying to contact patients, opt-out testing of people admitted to emergency departments provides an opportunity for patients to be re-linked to care.

Given the seriousness of the problem, with disengaged patients often only coming back to care when very ill (and costly to treat), Dr Childs called for more attention to be given to the issue. She suggested a national “Welcome Back” campaign.


This report is based on the following non-abstract presentations and accompanying discussions at the BHIVA 2023 Spring Conference in Gateshead.

Sullivan A. Overview of the cascade: where should we focus next?

Childs K. People living with HIV not in care: time to act!

Mower R. Can we identify patients at risk of not being in care?