FAO: HIV/STI Department, Health Protection Agency
Dear Sir or Madam,
HIV Pharmacy Association feedback to the proposed HIV & AIDS Reporting System (HARS) Dataset.
We would be grateful if you would consider our feedback on the proposed HARS dataset and in particular the Treatment Information section. I have listed the feedback under each Process ID and Field Name below:
Under Treatment Information:
42 ARVcode
90Blinded Treatment in a clinical trial - it is not clear from this description whether the patient is taking a blinded antiretroviral (ARV) treatment. Some trials are for non-ARV drugs. Should this not be split into two separate codes? Is this on the assumption that patient’s ARV treatment is fully funded while on the trial? If so, would it better to describe those trials which are funded and those that are not?
There is no Unblinded Treatment (or open label) code – e.g. PIVOT trial.
43ARVband
A First ARV regimen
B Second and subsequent ARV regimens
C Complex ARV
X Not on ARV
“Patients should only be categorised as C if they have: multi-drug resistance, multi-drug intolerance, significant co-morbidities, or significant drug interactions. A patient should not be placed in Category C if they are able to successfully take a guideline recommended standard ART regimen. This would place them in Category B despite having a complex ARV history per se. These combinations should be contructed following PEER review.
Q1. What is described as multi-drug resistance (is this 2 or more drugs (e.g. lamivudine and efavirenz), 2 or more classes of drugs (NRTI and NNRTI including etravrine and rilpivirine).
Q2. How many drug intolerances would count as multi-drug intolerance (2 or more drugs? (e.g. nevirapine and efavirenz)) or (2 or more drugs from different classes (e.g. efavirenz and atazanavir), or 3 or more?
Q3. What is classed as significant co-morbidity in the context of defining complexity? Is this a co-morbidity that prevents the use of first-line or second-line drugs due to a relative contra-indication or where there is a drug-drug interaction that dictates safer prescribing with a high-cost drug? (e.g. raltegravir use in patient receiving chemotherapy for a neoplasia)
Q4: What is classed as a significant drug-drug interaction? All patients receiving a boosted-protease inhibitor are benefitting from a significant drug-drug interaction between ritonavir and the protease inhibitor (PI). Many of these patients are on concomitant medication some of which will interact to an unknown extent with ritonavir and/or the PI. When does this become significant and how can you define this?
Case 1 example for classification:
If a patient on a boosted –PI attends clinic and is found to have been prescribed fluticasone inhaler by their GP which can lead to Cushing’s syndrome within a month do we record this at this visit as C complex, or if they are on Truvada/Atazanavir/ritonavir (which is a guideline regimen) is it now category B, yet we may have had to contact the GP, taken a complete drug history with medication review resulting in a letter to the GP.
Case 2 example for classification:
A 50 year old lady with a history of myocardial infarction, high blood pressure , depression and anxiety is found to be HIV+ve in clinic with a CD4 of 100 and a viral load of 136,000 copies/ml. She has fully sensitive virus. She takes amlodipine 10mg od, aspirin 75mg od, atorvastatin 40mg nocte, bendroflumethiazide 2.5mg od, bisprolol 5mg od, citalopram 20mg od, spironolactone 25mg od, telmisartan 80mg od. Her consultant, hears her concerns about any drugs that cause dizziness (relative contraindication for efavirenz), she also does not want to risk a yellowing of the skin or whites of eyes (relative contra-indication for atazanavir) therefore the consultant prescribes Truvada 1 tab od, with Darunavir/r 800/100 od. The pharmacist writes to the GP, detailing 3 co-prescribed medications that need dose modification to prevent toxicity and one medication that requires closer monitoring. In summary, we have a patient who is started on their first ARV regimen with significant co-morbidity, significant drug-drug interactions, and potential drug intolerances. Which category of ARVband is appropriate at this visit?
A First ARV regimen
B Second and subsequent ARV regimens
C Complex ARV
We feel this patient would be correctly described as C complex, however in the explanatory notes B should be used, however A may be appropriate as this is the first regimen.
It seems unclear what the classification here is aiming to define. There will be different interpretations of the patient case and we may find that many patients end up being classified as B even though they have intolerance, resistance, drug-drug interactions and significant co-morbidity. This does not reflect the support and complexity around managing these patients.
44Homedelivery
We would recommend calling this Community Supply. Some patients receive their medication at home by delivery, many by postal service, and now many more by collecting at a community-based pharmacy. Is it not the case, that what is being asked here is whether the supplies that the patient receives are VAT-free? In which case we would recommend that the field is described at Community Supply.
The categorical “Yes/No” response here is clear however “No” does not tell us whether that patient has been approached and declined. Knowing the potential uptake in cohorts is helpful we believe and also a saturable process that may never reach 100% where patients have serious concerns about confidentiality and wish to have their medicines dispensed by the hospital-based pharmacy. For those who do not have stable disease hospital dispensing is an important
We would propose two further categories against this field:
“Yes, No, Declined, Not eligible”
45Clinical Trial Indicator
The episode of care for a patient is part of a clinical trial. Again, in reference to our comments in relation to field 42, is it important to establish here is this episode of care funded through a clinical trial, and is it related to ARV therapy or not?
Adherence
We have learnt that patients on treatment with Viral loads >50 copies/ml will be counted as low adherers. HIVPA are particulary concerned about this definition and way of gathering adherence information. Using viral load as a surrogate marker is better at telling whether a patient is adherent than non-adherent. There is no evidence that viral loads in the 40 to 400 copies/ml range are associated with low adherence. Some viral load tests are particularly sensitive and tend to produce detectable viral loads where as older tests can reveal an undetectable result (e.g. Roche Taqmanvs Chiron bDNA).
Drug-drug interactions and drug resistance may also be common reasons for detectable viral loads, therefore the assumption that all detectable viral loads are due to low adherence places the patient as responsible where in fact it may be due to a primary care prescribed drug that interacts (e.g. omeprazole with boosted atazanavir) or undetected drug resistance (e.g. adherent patient taking a failing regimen).
The reason for collecting this information is not clear and there are many causes of low adherence, particularly psycho-social issues which appear not to have been included as signifying complexity. Many patients who are suspected to be low adherent are given extensive support in clinic with frequent visits and this is why we can maintain them with undetectable viral loads, however they would not be identified in this measurement. We do not believe that low-level viraemialess than 100 copies/ml with Roche Taqman should necessarily prompt an escalation of support without an adherence assessment and a case-by-case review with an expert advisory group of clinicians including a virologist.
And finally, we feel that there is sufficient evidence to demonstrate that some vaccinations cause low-level viraemia that may persist for several months that can be misread as low adherence. This further underlines the inappropriateness of using low-level detectable viraemia as a surrogate marker for low adherence.
We would like the agencies to respond to each of our questions and consider our response as we believe that the data collection under Treatment Information can be improved, made more robust and provide more accurate data from which, conclusions may be drawn.
We would like to thank the Health Protection Agency, the National Reference Group for PbR and the Department of Health for allowing us this opportunity to comment on this important revision of datasets that impact on clinical services. The HIV Pharmacy Association is committed to providing expert advice on medicines optimisation in the HIV specialty. We are happy to advise and support National work on gathering treatment information therefore please do not hesitate to contact HIVPA if you have feel we can offer support to this initiative.
Yours sincerely
David Ogden
HIV Specialist Pharmacist
On behalf of the HIV Pharmacy Association