Alex Nowbar’s research reviews—10 September 2018

This week we introduce the next of our three new columnists who will be writing our weekly research reviews

Alex Nowbar is a clinical research fellow at Imperial College London

Reducing harm from cigarettes—a serious strategy

This expert psychology team are thinking big when it comes to reducing harm from cigarette smoking. They conducted a large double blind randomised controlled trial of reasonable duration (20 weeks) comparing immediate nicotine reduction to gradual reduction, with no reduction as the control. The primary outcome was biomarkers of smoke toxicant exposure. When there is so much evidence for people to simply stop smoking as soon as possible, you may wonder why anyone would conduct such a trial. But this research is targeted beyond individuals who smoke. It is about changing the content of all cigarettes being sold to make them less addictive. They found lower toxicant exposure in the immediate reduction group compared to the gradual group and no differences between the gradual and control group. The other benefits of immediate reduction included reduced dependence and more cigarette-free days. The downsides of immediate reduction were also fairly predictable—more withdrawal symptoms, more use of non-study cigarettes and higher drop-out rates. But if people had no access to products with more nicotine (by enforcing low nicotine content on all combusted products), the true benefits of a low nicotine strategy could quickly be realised.

Medicare joint replacement bundles—the incentive impact

Today I forgot to put on my economics helmet and came across these exciting-sounding things known as BPCI and LEJR in this paper about a Medicare program. After much study, I have decided “bundled payments for care improvement” simply relate to how hospitals get paid for the work they do. In the UK I think we use coding to calculate such things. As for “lower extremity joint replacement episode,” I concluded they were pulling my leg. And yet, despite all the unfamiliar acronyms this study was highly worthwhile. It assessed for potential undesirable effects of a payment reform. The payment structure moved away from traditional reimbursement per activity such as tests, drugs, and procedures and instead focuses on provision of care for a whole admission based on the admission diagnosis, in this case joint replacements. The idea was to incentivise shorter stays and to deter ordering of unnecessary tests, but it could also have led to doing more procedures when not clinically indicated and doing those procedures in only the lower risk patients. The study’s findings were generally “reassuring” in this regard, but there was one case-mix factor (out of the 20 examined) that was different in hospitals participating in BPCI; patients who had previously been at a nursing home were less likely to undergo a joint replacement. We can never know if this observation is actually a result of the payment reform, but even if it was, it’s hard to say whether this change in case-mix is good or bad.

Removal of QOF financial incentives—the quashed quality

Back in familiar territory, but no further from the acronyms, Minchin et al wanted to know the impact of removing the NHS’s favourite outcome-driven financial incentive. The study showed an immediate decline in performance on quality measures. If I were an optimist I’d say this doesn’t mean patient care suffered because maybe it was only the incentive to document that was removed. But I am not an optimist and financial incentives are powerful things. Just remember that with great power comes great responsibility.

Capping influenza—a viral victory

Flu season is almost upon us. Or should I call it reluctant oseltamivir prescribing season. Thankfully, there are some promising data on a new antiviral, baloxavir, which acts on a different part of the virus to oseltamivir. Hayden et al report two trials, one phase 2 and one phase 3, called CAPSTONE-1. In CAPSTONE-1, a double blinded placebo-controlled trial of over a thousand healthy-sounding patients, the time to alleviation of symptoms was significantly shorter with baloxavir. 20.7% of patients in the baloxavir group experienced an adverse event. This figure was 24.6% and 24.8% in the placebo and oseltamivir groups respectively. Another bonus is that baloxavir is given once only (oseltamivir is twice a day for five days). Their next trial tests baloxavir in patients at higher risk of complications, those whom guidelines say we should give oseltamivir to currently, but it might be hard to detect a benefit since both study arms will also receive oseltamivir.

CT coronary angiography and cardiovascular events—the irradiation intrigue

Tests of tests can be pretty dull. Sensitivity and specificity competitions aren’t much fun. The SCOT-HEART investigators took an exotic approach with CT coronary angiography for stable angina. They compared addition of CT coronary angiography to the patient care pathway to standard care in an open-label randomised controlled trial. The primary endpoint was death from coronary heart disease or nonfatal myocardial infarction at five years. Reduction in these events is a bold aim for a test (and was not their originally declared aim, which was more about diagnosis), but apparently it succeeded, with a hazard ratio of 0.59. I (and NICE) already thought CT coronary angiography was a useful test, but I never expected it to directly improve cardiovascular outcomes. My expectations were modest—maybe shorter waiting times, fewer unnecessary procedures, and lower overall costs. How can an imaging test bring that much cardiovascular benefit? One possible explanation is that the CT group got more preventative drugs like aspirin and statins. Another explanation is that unblinded event measurement biased the results in favour of CT.

More CT coronary angiography—a fatty future

On the subject of CT coronary angiography, the CRISP-CT study showcased a new imaging biomarker called perivascular fat attenuation index. The idea is that this index indicates coronary inflammation which might be more relevant to future events than a traditional measure like calcium score. And sure enough, using attractively large derivation and validation cohorts, they found their biomarker was predictive of all-cause mortality and cardiac mortality. And thus I can safely emit the standard spiel that this may be valuable for targeting preventive therapies at high risk patients. But I wouldn’t make any claims beyond that.

Cardiovascular risk of NSAIDs—a confounder challenge

NSAIDs are commonly used by patients who have an inflammatory condition. Inflammatory conditions are associated with cardiovascular risk. This confounder makes it difficult to know if an NSAID increases cardiovascular risk without a randomised controlled trial. Of course, such a trial is not feasible as it would require too many patients and too long a follow-up period. Schmidt et al’s study, covering millions of Danish registry patients from 1996 to 2016, may be the world’s best attempt at addressing this question. Five groups of patients were compared, those starting diclofenac, ibuprofen, naproxen, or paracetamol and non-users. The results are striking; major adverse cardiovascular events were 20% more common in diclofenac initiators than in ibuprofen initiators, 30% more than in naproxen initiators and 50% more than in non-initiators. Diclofenac also came out worse than paracetamol for cardiovascular events. Gastrointestinal bleeding rates were highest for diclofenac and naproxen. So, does the bell toll for diclofenac? Not quite. These data still have the limitation of being observational despite the smart methodology used. The absolute risk is still relatively low. And patients may still prefer diclofenac over the others because it works (but we won’t know unless we ask them).

Fish oil in pregnancy—a supplement success

The Copenhagen Studies on Asthma (COPSAC) 2010 birth cohort contained a nested randomised controlled trial of a fish oil supplement (an n-3 fatty acid) or olive oil taken from week 24 in pregnancy until one week postpartum. This trial was double-blinded and the primary outcome was persistent wheeze or asthma. This week in The BMJ, the authors report on one of their secondary outcomes, growth. At six years old, BMI z score (which adjusts BMI for child age and sex) was higher in the fish oil group. They also diligently measured body composition using dual entry x-ray absorptiometry (DEXA) and again found the supplement to be beneficial. Crucially, they stipulate that the improved growth did not lead to an increased risk of obesity. However, the cohort continue to be followed-up so it may be too early to draw conclusions. These findings are encouraging and, according to the trial registration, there is plenty more data to look forward to, including neurological development and eczema.

Statins in the old and very old—a nonagenarian negative

Here is some evidence to inform practice for octa and nonagenarians, a rare form of evidence indeed. Ramos et al’s large retrospective cohort study looked at the impact of statin use on atherosclerotic cardiovascular disease and all-cause mortality in people over 75 without pre-existing atherosclerotic cardiovascular disease. They divided people into those with type 2 diabetes and those without. They found no benefit of statins in those without diabetes. There was benefit in those with diabetes, but this benefit was less over age 85 and disappeared in those over 90. A randomised controlled trial of statins in over 70s (STAREE) is underway so we should soon have a definitive answer which takes adverse events into account. This will be a great step in efforts to tackle polypharmacy in the elderly.

Build urgent care centres and they will come—a debatable diversion

Emergency physicians in the US may be pleased to hear that attempts to divert people who are less sick (“low-acuity patients” with ailments like rashes and musculoskeletal strains) away from emergency departments seem to be working. Emergency department visits decreased from 89 visits per 1000 members of the Aetna healthplan (the cohort studied here) to 57 per 1000. Far more marked though, was the increase in non-emergency department visits, which rose from 54 visits per 1000 to 131 per 1000. And with that, Poon et al showed a hefty increase in utilisation of resources and in costs. Non-emergency department visits were mostly comprised of urgent care centre visits, but the authors also show the growing popularity of “retail clinics” and “telemedicine,” hardly surprising with out of pocket costs (paid by patients not insurers) of $37 dollars and $14 per appointment respectively compared to $422 at the emergency department. Some of this study is lost in translation over the Atlantic though, because we have another arrow in our quiver, that is, rapid (and free) access to general practice. Still, these data highlight the unmet needs of a population which is not that different to the UK’s. Uptake of these types of services is definitely growing. Unfortunately the analysis only covered those under 65 which holds limited relevance to the bigger challenges faced by healthcare systems today. 

Alex Nowbar is a clinical research fellow at Imperial College London.

Competing interests: None declared