Some London clinics may have difficulty meeting drug prescription targets

Some HIV clinics in London may have difficulty conforming to the new targets set for the proportion of patients who are prescribed cost-saving regimens, according to a poster presentation at this month’s BHIVA conference.

The poster, from Guy’s and St Thomas’s Hospital Trust, shows that despite there being a considerable financial incentive involved, the HIV clinic at St Thomas’s Hospital failed to achieve a target of putting at least 80% of new patients on non-nucleoside (NNRTI)-based regimens instead of the more expensive protease inhibitor (PI)-based ones, and documents the reasons.

Background – prescribing changes

The question of which drugs patients are preferentially prescribed became more prominent this month when the London HIV Consortium of the London Specialised Commissioning Group (LSCG) announced that it had agreed with London’s lead HIV consultants that they would prescribe abacavir/3TC (Kivexa) plus efavirenz (Sustiva) as their regimen of first choice for drug-naïve patients, and boosted atazanavir (Reyataz plus Norvir) as their first choice of PI, in preference to more expensive alternatives.

The LSCG has now placed an explanation and presentation on its new prescribing policy on its website – see here.

This has been done as a way of saving approximately £9 million which would otherwise have to be saved by reducing services. London has received no increase in its HIV treatment and care budget this year despite a projection that the number of patients on antiretroviral drugs will grow by at least 5.3%.

Claire Foreman, lead commissioner of the London HIV Consortium, told its Patient and Public Engagement group today that the Consortium pledged that it would not use the ultimate sanction of withholding a Trust’s HIV drug funding if it failed to conform to the new drug prescribing requirements.

Short of this sanction, however, the NHS already has a number of financial rewards schemes designed to induce individual trusts to reach specific targets. One of these is the CQUIN (Commissioning for Quality and Innovation) payment framework, under which trusts get a proportion of their income on condition that they meet certain targets and are penalised if they do not.

The ‘80% NNRTI’ CQUIN was initiated in the financial year 2009-10 by most London HIV clinics. Prescribing NNRTIs rather than PIs saves, on average, £1500 per year per patient. The 80% target also includes patients that have NNRTI resistance, but excludes patients who cannot take NNRTIs for other reasons, such as pregnant women (who cannot take efavirenz) who have high CD4 counts (and so cannot take nevirapine). These are supposed to account for the other 20% of patients.

The Guy’s and St Thomas’s Study

Despite the considerable financial incentives, Guy’s and St Thomas’s only ended up prescribing NNRTIs for 50 out of 80 new patients in the year December 2009 to November 2010. There were another five patients who were excluded from NNRTIs because they had resistance. This meant that only 68.75% of patients met the CQUIN target instead of 80%.

There was no difference in CD4 count between patients given NNRTIs (218 cells/mm³) or not (223 cells/mm³), but four people in the NNRTI-sparing group had AIDS-defining conditions versus one in the NNRTI group.

Of the 25 patients who were offered NNRTI-sparing regimens despite having no drug resistance:

• Six patients were deemed ‘chaotic’ and so were offered PI-based regimens, which are more robust in circumstances of poor adherence;

• Three were women planning pregnancy who had CD4 counts too high for nevirapine;

• Three had working patterns that precluded efavirenz (which is usually taken at night) but could not take nevirapine because their CD4 counts were too high or they had hepatitis B (nevirapine can cause liver failure);

• One was on drugs that interacted with NNRTIs.

This left nine patients who were offered NNRTI-sparing regiments because their drug resistance test had not come back. If clinicians had waited and none of these patients had had resistance, then prescribing NNRTIs would have meant hitting the 80% CQUIN target exactly.

Lead author Selina Singh, however, told aidsmap that there were good clinical reasons for prescribing PIs for most of these nine patients too.

Four patients had AIDS-defining illnesses and their physicians did not feel that delaying ARV treatment was safe.

Two had abnormal liver function; one also had a rash and the other was extremely worried about psychological side effects; in both cases physicians were therefore concerned about NNRTI-related side effects.  

Two patients with low CD4 counts were established on PI-based regimens and there was a delay in sending resistance test results. By this time the patients were established on their regimens and did not wish to switch.

This left one patient who wanted to be on a PI-based regimen because he wanted the same treatment as his partner.

In the vast majority of cases, then, physicians prescribed NNRTI-sparing regimens because they believed there was a clinical need for them.

The authors comment: “Those starting on a PI were more likely to have an AIDS-defining illness. In such acutely-unwell patients, it is often safer to start on a PI than risk [liver] toxicity.”

“This indicates that CQUIN targets are unrealistic for a cohort of patients needing to start antiretroviral therapy,” they conclude.