Audit of treating naive patients showed basic UK guidelines not followed in over 25% patients; 60% patients start with a CD4 count <200 cells/mm3
Simon Collins, HIV i-Base
The annual BHIVA audits are one of the most important sources of information on current standards of HIV care in the UK.
Each audit is usually linked to a BHIVA clinical guideline or a particular aspect of care. This year was the first prospective audit. The subject was the initial treatment of naive patients, and results were presented by Emma Street at the Autumn conference.
Key conclusions from the audit included:
- 4% of patients were lost to follow-up within the first year.
- 60% patients started treatment at a CD4 count that was less than 200, and that starting at the lowest CD4 counts was associated with poorer outcome.
- 1 in 6 patients still had a detectable viral load 5-12 months after starting treatment.
- At least 16% patients did not have a resistance test prior to starting treatment.
- 54% patients did not have a viral load result by week 4 and 26% did not have a result by week 6.
- 25% patients with hepatitis B coinfection were using 3TC without tenofovir (HBV monotherapy).
- There were no apparent differences between results at larger and smaller treatment centres.
Results were collected on 1301 treatment naive patients from 133 centres who started first-line therapy from April to September 2006. Unfortunately, 11 centres did not take part in the follow up phase, accounting for 86 patients. Major discrepancies in data excluded a further 45 patients. This left 1170 patients included in most analyses. The primary outcome was the viral load result closest to six months after starting treatment.
Baseline demographics included 54% men, 44% women (some not stated); almost 50% African, 38% White, 4% Caribbean; and 16% were pregnant women.
CD4 count when starting treatment, showed that the majority of patients already miss the minimum stage recommended for an optimal response to treatment: 19% had a count between 0 and 50, 41% were between 50 and 200, 30% started between 200 and 350. Approximately 5% started between 350 and 500, and 5% at over 500 cells/mm3.
Around 70% of those starting <200 cells/mm3 had been diagnosed within the previous 6 months (ie late diagnoses), but over 6% of those who had been diagnosed for at least 6 months still started at less than 50 cells/mm3. This probably reflects a small percentage of patients who are still extremely reluctant to start treatment. Baseline viral load was greater than 100,000 copies/mL in 43% patients.
Only 72% patients had resistance test results available prior to starting treatment, with 6% still waiting for results and at least 16% not having been tested. Of those tested 6% had evidence of single class resistance and 1% had multi-class resistance a prevalence that is high enough to justify the current recommendations for all patients to be tested on diagnosis or prior to starting treatment.
Approximately 35% of patients each used either tenofovir/FTC or abacavir/3TC as dual nukes, but 25% of patients still started treatment with Combivir. Even if all the women using treatment during pregnancy started with Combivir, this leaves at least another 10% of patients were prescribed this non-recommended option.
55% patients started an efavirenz-based combination, with just under another 20% using either nevirapine or lopinavir/r. Approximately 5% patients started with either saquinavir/r or fosamprenavir/r.
The primary outcome of suppression to <50 copies/mL was achieved by 68% of 1215 patients. 14% were detectable (and not accounted for by 6% who stopped treatment used during pregnancy). 12% had no outcome reported. In an on-treatment analysis (patients still on treatment with a measurement at 6 months), 84% were suppressed to <50 copies/mL and 16% had detectable viral load (approximately 65% of these patients were suppressed to 50-500 copies/mL).
The percentages of patients suppressing to <50, 50-500 and >500 copies/mL were similar across all clinics (from those treating less than 50 patients to those treating more than 500 patients).
Time to first viral load result was also a concern, with only 46% patients having a viral load result by the recommended maximum of 4 weeks after starting treatment. 75% of patients had a result by 6 weeks and 85% by 8 weeks.
Given the importance of confirming viral activity, and the opportunity to provide both adherence support and evaluation of side effects, this appears to represent one of the more alarming findings,
Results showed that approximately 25% patients with either baseline CD4 <50 cells/mm3 or viral load >100,000 copies/mL did not reach than 50 cells/mm3.
Of the 151 patients who had no primary outcome result, 66 had transferred care, and a similar number were neither known to have transferred care or were untraceable a poor outcome for patients starting first treatment. 11 patients died. Patients stopping treatment and not transferring care were disproportionately more likely to be African (72%) and female (60%), compared to the group as a whole.
The contribution of dispersal and being denied leave to remain in immigration cases was unclear in these results.
Several other secondary outcomes were briefly reported:
- On adherence: ‘no known issues’ were report for 85% patients a term that could cover healthcare indifference and lack of knowledge with 10% reporting ‘some problems’ and <5% ‘substantial problems’.
- Hepatitis B coinfection was reported in 53 (4.5%) patients, 13 of who were using 3TC without tenofovir. Just over 6% of patients were reported as being untested for HbsAg (although some clinics may have used HbcAb, which was not included in the survey).
- 2.7% patients were hepatitis C antibody positive at baseline and 4.8% were untested, despite this also being a clear recommendation for routine standard on care on diagnosis.
Some of these results are disappointing and it is therefore all the more important that this audit was commissioned and that the results were presented at this national meeting.
This audit seems to indicate that despite clear guidelines, suboptimal treatment is being delivered to a significant proportion of patients.
The results provide a snapshot of strengths and weaknesses but are unable to give much detail about the reasons that could be behind some results.
Early CD4 and viral load tests at 2-4 weeks have long been established as standard of care, and additionally provide the opportunity to monitor for side effects and support adherence. Some patients appear to be given three months treatment and told to come back when that is finished unless there is a problem. This is unacceptable.
It is easy to understand why someone would not start without a CD4 count, so why is resistance testing still a hurdle? Cost is not an excuse it is probably covered by delaying starting treatment by a week or so (and yes, we know about different budgets).
It is disappointing that some clinics who took part in the first part of the audit did not produce follow up results. This, together with the missing and loss-to-follow-up data probably overestimates the success of treatment that was reported. Although the audit produces less rigourous data than would be available from a clinical trial, BHIVA should be commended for highlighting this picture of current clinical care.
These results should focus and inform clinicians, advocates, patients and purchasers alike.