Lost to care: the mystery of the disappearing patients

Up to a third of patients who attend their first HIV clinic appointment don’t return. Why do they drop out of care, and are they putting themselves in danger? Chris Morley, co-ordinator of HIV policy, information and publications at George House Trust in Manchester, and HTU’s editor Gus Cairns investigate.

In the last issue of HIV Treatment Update we looked at ways to increase HIV testing in the UK and minimise the number of people who don’t know their HIV status. However there is not a lot of point in people learning their positive status if they then simply disappear from care.

How much does this happen, and does it matter? We already know that patients who are diagnosed after their CD4 count has fallen below 200 cells/mm³, are eleven times more likely to die in the immediate post-test period than patients diagnosed sooner.¹

If figures from France are anything to go by, this matters a lot. A study at five HIV clinics near Calais² found that over one in eight patients (13%) dropped out of care for at least a year.

Over half of those subsequently returned, after an average gap of 19 months. Nearly half of them by that time had a CD4 count under 200 cells/mm³ and a quarter reappeared with an AIDS-related illness. And after the researchers controlled for the influence of patients’ CD4 counts at diagnosis, they found that dropping out of care was associated with a more-than-fivefold increase in the chances of dying in the year after return, compared with people who had stayed in care.

We already know, from figures from the US, that 30% of people who take a rapid HIV test and get a reactive (positive) result fail to return for the necessary confirmatory test.³

London dropouts

At least as high a proportion of patients in the UK as in the US seem to be disappearing from care. Last year, a study from King’s College Hospital in south London⁴ found that no less than 40% of patients seen at least once between 1995 and 2005 were not seen at all during 2006. Checking with the Health Protection Agency (HPA) found that half of these were attending another clinic and a small number were known to have died, but it still meant that more than one in five of all patients had disappeared from care. (The HPA encodes patients’ surnames with a method called Soundex which, when combined with date of birth, creates a distinctive, though not necessarily unique, identifier which can be followed from clinic to clinic. No other identifying detail is kept.)

This year, several surveys from clinics in London were presented at the BHIVA conference in Liverpool with similar findings. One was from Homerton Hospital in Hackney, east London. This found that it was not patients failing to turn up for confirmatory tests that was the problem; instead, they were dropping out after this, once they’d had their confirmatory test and their first serious chat with an HIV consultant.

Only one out of 88 newly diagnosed patients failed to turn up for their initial HIV-clinic appointment after they had been diagnosed in other settings ranging from the GUM clinic to antenatal clinics, and by their GP. After that, however, over a third of the total failed to turn up for any subsequent appointment during 2007, including a quarter of the patients with CD4 counts under 200 cells/mm³.

Dr Iain Reeves from Homerton comments:

“We can’t eliminate absolutely everyone who turned up at another clinic, as the Health Protection Agency data we use is very sensitive to misspelt names, wrong dates of birth, and so on.”

Nonetheless, it’s an ongoing problem, and has continued in 2009, he adds. If you take a shorter time-frame, you get better attendance, as cases are ‘live’ and exclude people who have died or moved abroad.

“But,” comments Reeves, “in our clinic population of about 700 we had 56 (8%) who we’d seen in the second half of 2008 who we didn’t see in the first half of 2009. We found that ten of these had transferred to another clinic and five had died and we’ve managed to get about eight to come back by phoning them, but that leaves 33 patients who have completely vanished – nearly one in 20.

“If we’ve got a phone number or address, we’ll try and contact them and if we’ve got permission to contact their GP, we will do that. In rare circumstances, if someone has been really seriously unwell, we may break confidentiality and contact their GP anyway, though I can only recall a couple of cases. There is a dilemma between respecting patient autonomy and our duty to ensure the patient doesn’t become seriously ill or die.

Community organisations often spend time unpicking and correcting inaccurate or incomplete messages seared into people’s brains in the trauma of diagnosis.

“People may have all sorts of reasons for not turning up. One of my patients was in prison and turned up again once he was out. But the two biggest reasons are depression and fear of stigma. There is a subset of patients who struggle with depression, have problems with adherence, and are very difficult to get into the clinic. With one recent patient of mine we counted it as a success that she’d phoned up to cancel her last appointment instead of just failing to show.

“Another lady was admitted here very ill and went straight to the Intensive Care Unit where she unfortunately died.  Her partner tested positive here and then he disappeared too. When contacted he said he had seen some people in the waiting room he thought might know him and he didn’t want to be identified.”

The North Middlesex Hospital, just a few miles away from Homerton, also did a survey, this time of long-term rather than new patients, and found that between 2005 and 2009, about 6% of the patient population disappeared every year. This figure excluded those known to have transferred, died, or moved out of the UK. It managed to trace 44% of those missing but could not establish where the remainder had gone.

Patients lost to care were on average somewhat younger, more recently diagnosed (two versus five years) and, worryingly, much more likely to have a detectable HIV viral load (60% versus 20%). Three times as many were on a failing drug regimen at the time they had disappeared from care as patients who stayed in care.

The North Middlesex is now in the process of a systematic attempt to trace these patients. The hospital’s Dr Chris Wood explains:

“We find the best results are to phone people’s mobiles, as in our clinic population, many of whom are immigrants, their mobile numbers change less than their addresses. So far, we’ve managed to persuade everyone to re-attend that we’ve actually spoken to.

“Many are women diagnosed in the antenatal clinic. One recently phoned up and said she thought she had an appointment in November- after 2.5 years! I think she was embarrassed about not seeing us.

“Women sometimes prioritise their children’s needs over their own, and may only show up again when they become pregnant once more: the way we imply that the purpose of testing is to avoid transmission to the baby may reinforce this.

“A lot test positive out of the blue, and feel perfectly well; in some ways they’ve never quite believed the test result. Some are in wilful denial, some genuinely don’t know if they need to stay in touch, and some have very controlling spouses who don’t want them to be in care.

“Another common reason is the dispersal of immigrants. Patients who don’t make links with a clinic in their new area they will drop out of care. It may have taken them a long time to trust a healthcare team and they don’t want to go through it again.

“And yes, some believe they are cured. I had an east African chap who insisted he was. He said ‘I know you won’t agree with me, but I think I’m cured. I want a viral load test to prove it.’ Well, we did the viral load test, which was high, and he still wasn’t impressed by it!”

Missing in Manchester

This is not a problem restricted to London. The HPA has confirmed that over 300 people living with HIV in north west England are missing from UK HIV clinics.

Around five or six years ago George House Trust in Manchester, the region’s leading HIV community organisation, noticed that the number of new HIV cases was never matched by a similar rise in the total number of people documented as living with HIV: there was always an excess of new cases. Close study of the annual HIV reports⁵ seemed to show NW England was steadily leaking people with HIV, despite a rising population.

As long as only a few people went missing, this could be plausibly explained – for example, some people in Cumbria in the north of the region find it easier to go to Newcastle-upon-Tyne, and therefore out of area, for care.

But in 2007 the gap between new cases and the rise in the region's HIV total widened to more than twice the size it was before. George House Trust was concerned and wanted some answers. The staff there did a detailed analysis tracking the loss for each year since 1996 and then discussed their results⁶ with the Centre for Public Health at Liverpool John Moores University, which produces the region's HIV statistics. John Moores confirmed that while slightly over 100 were now attending clinics outside north west England, there were more than 300 NW residents entirely missing from UK HIV clinics – that’s 5% of the total diagnosed, as in London.

But patients lost by hospitals may not have disappeared from the community. Near the back is an interesting table that reveals that a third of the 2834 people living with HIV who used a community sector service in 2008 – a facility like George House Trust - were not in contact with any of the region's HIV clinics that year, and that fully one-quarter of people using community services have never been seen since diagnosis by a clinic in the region.

That represents an opportunity for community services to talk to their service users and find out if they are among the missing patients, and if so why and what they feel their needs are.

People may have all sorts of reasons for not turning up. One of my patients was in prison and turned up again once he was out. But the two biggest reasons are depression and fear of stigma. Dr Iain Reeves, Homerton Hospital

For instance, we can see that among the nearly 300 asylum seekers using community services in 2008, a third had never been seen by a clinic since diagnosis and 40% made no contact during 2008. UK nationals do somewhat better, but just under 30% made no contact. People from any minority ethnic group are more likely to be missing from clinics than white people, but 29% of white people had no clinic contact either.

Over a quarter – 27% - of gay men who used community sector services were unseen by clinics in 2008 and more than half of these men have never been to a NW region HIV clinic since diagnosis.

Finally, a startling 60% of children and young people with HIV who are in touch with community services – that’s 40 out of 67 young people - have never been seen since diagnosis by a clinic in the region. Yet both Manchester and Liverpool have specialist regional paediatric hospitals.

People who’ve never used a NW England HIV clinic since diagnosis are more likely to be male than female, and heterosexually infected rather than gay men (53% compared with 39%). Conversely, however, people who had not turned up in 2008, having been seen in the previous year, were more likely to be gay white UK nationals.

This apparent contradiction can be explained if vulnerable and marginalised people (especially women diagnosed in pregnancy and young people) are more likely to drop out of care after diagnosis altogether, and are only forced back by sickness, but good CD4 counts result in people deciding to take a break from care – hence the shorter-term absence of gay men.

We are doing better than we were. HPA research suggests being lost to follow-up is a stage many people go through and that the proportion of clinic drop-outs was once much higher, especially before effective HIV therapy became available. The estimated drop-out rate plunged from 14% before combination therapy to 6.3% in 1998, and then gradually to 3.8% in 2005.⁷ Almost everyone appears (or reappears) for HIV clinical care at some stage. In London, Chris Wood comments that “people often show up after a gap of three to four years.”

Why do people drop out?

Tables and studies can’t tell us anything about the reasons people don’t use clinics. One of us (Chris) talked to several people at George House Trust who identified themselves as former or current drop-outs.

Here’s an example:

“When I was diagnosed back in ‘93 there was still a 'death sentence' attitude both from clinical staff and other positive people I knew. I remember being told by the nurse to make the most of the time I had left, and the doctor wrote me a DS1500 [people expected to die within about 6 months can use this to claim Disability Living Allowance immediately]. It was another six years before I went back to clinic, after having somewhat of a wild time in between. I realised I hadn't kicked the proverbial [bucket = died], but was starting to notice higher levels of fatigue, which prompted me to go back.” (white gay man, mid 40s)

Talking with people living with HIV reveals a wide range of reasons why people become lost to follow-up. People often have several reasons and these may shift over time. One man who was diagnosed this May, has an aversion to needles and lacks suitable veins for drawing blood, and has now given up on the clinic because he feels he’s “been roughly handled and not given respect and privacy”. He’s also “depressed and anxious and [doesn’t] want to stress about my CD4.”

Few people see HIV as a purely clinical issue in their lives. For some people no amount of clinical good advice and reasoning will work because this simply doesn’t fit their frame of reference for understanding and living life. For some people, for example, unless drugs and their taking are explained and legitimised in Bible terms (by reference to a verse or short phrase), they may not take treatment. Unless we understand and engage with people’s individual socio-cultural beliefs, the chances of engaging and retaining the person in clinical care are diminished.

There are other reasons for drop-out, such as impractical clinic hours not geared to those with jobs:

“I book an early appointment to minimise time off from work but am always made to wait at least an hour; pharmacy home deliveries don’t arrive as booked although I have taken time off for these; and I can’t give bloods for convenience at my GP. The service has a ‘take it or leave it’ attitude, one size fits all.”

People clutch at bits of HIV good news and are ignorant about or disregard much else they may hear. In the midst of the shock of diagnosis, there is evidence of selective listening and over-simplistic reassurance, and both contribute to drop-outs.

People gave us lots of examples of what might be called over-enthusiastic reassurance. We’ve heard: “You are more likely to be knocked down by a bus than die of HIV,” “You’ll probably live to 95,” “Your CD4 is 650 and you won’t need to start treatment for years,” and the classic “With modern treatments it’s just a long-term condition like diabetes.”

The problem with giving such reassurance is that in people who are already highly anxious about HIV it can become fuel for denial. Reassurance needs to be tailored to the individual. It’s often misplaced and misread, and taken as permission to put HIV aside and drop out of care.

Community organisations often spend time unpicking and correcting inaccurate or incomplete messages seared into people’s brains in the trauma of diagnosis. What people actually hear when HIV is compared with diabetes is: “You’ll live a more or less normal life”.

Ordinary people are unlikely to know that successful long-term diabetes care requires significant lifestyle changes in diet, weight, alcohol use, self-testing of blood-sugar levels and injecting insulin, and that if you don’t adhere to these you can lose your sight, have serious kidney problems, develop gangrene and need amputations. People aren’t told or don’t hear the full diabetes comparison story – that many people fail to achieve proper diabetes self-care and pay the health consequences.

When people are lost to follow-up with HIV, we need to make plain there will probably be a price to be paid in worsened health. Clinics need to be scrupulously accurate and fully explain what they are saying because people traumatised by bad news don’t absorb information reliably.

Chris Morley makes six recommendations to reduce the number of patients lost to care.

1. We need to better understand the reasons people drop out of care. These could include stigma, shame, fear, disbelief, misunderstandings about HIV and treatment, mental ill-health, adverse life circumstances, conflicts with the individual’s socio-cultural beliefs and not prioritising HIV in a busy life.
2. The HPA needs to improve its analysis of people who drop out and should feed back anonymised details to HIV clinics and community organisations.
3. Community research among people living with HIV is needed, to find out exactly why and for how long people drop out of care after diagnosis.
4. Organisations like the British HIV Association and the London HIV Consortium could audit HIV clinics for patient retention.
5. Good practice guides developed by organisations like BHIVA and further training can help clinics to predict likely drop-outs and meet their needs better.
6. Community organisations need to be aware that many of their service users may not be attending clinics and take steps to identify and better support them.