Alex Nowbar’s research reviews—30 April 2019

Alex Nowbar reviews the latest research from the top medical journals

Annals of Internal Medicine

The marmite of bone protection

Bisphosphonate therapy—you either love it or hate it. This systematic review of randomised controlled trials and observational studies of women with osteoporosis explains why. Four years of alendronate therapy reduced the risk of clinical fractures, as did six years of zoledronic acid. But with long-term bisphosphonate use there was also an increased risk of atypical femoral fractures and osteonecrosis of the jaw. The data available did not provide a clear answer for the use of drug holidays and drug discontinuation, but overall there was no worrying increase in clinical fractures in the discontinuation group versus continuation in the trials after five years of treatment.

Scam awareness and cognitive impairment

Scam awareness requires a high level of brain function. It encompasses a lot of thought processes that we tend to take for granted. I imagine good scam awareness is enhanced by exposure to personal or other people’s experiences of being scammed and the ability to learn from that. Boyle et al performed a cohort study of 935 older people without dementia at the start of the study to assess whether decreased scam awareness was associated with risk of developing Alzheimer’s disease in the future (including pathological findings from those participants who died). This well-executed American study found a strong association between low scam awareness and development of future cognitive impairment, even after adjustment for overall cognitive function. While a scam score is not about to become a clinical screening test for dementia, the questions in the scam questionnaire highlight important areas for strategies to prevent the elderly being defrauded.


Are you sitting comfortably?

According to Yang et al, you probably are. They looked at data on 51 896 individuals and found that 60% spent at least 2 hours a day sitting watching television or videos in 2015-2016. The proportion of people using a computer for at least 1 hour a day outside of school or work increased from 2001 to 2016 in all age groups. Total sitting time a day also increased. Interestingly, the prevalence of sedentary behaviour was higher in males in all ages even after multivariable adjustment.


Individualised nutritional support

The EFFORT trial was a large multicentre randomised controlled trial of nutrition support from dieticians initiated within 48 hours of admission to hospital. The control group had no dietetic input. The patient group assessed in this trial had to be considered at high nutritional risk through a score and also had to be expected to be in hospital for more than four days. There was a mortality benefit for those in the treatment arm as well as a host of benefits including improved quality of life. However, the authors rightly state “What cannot be measured in this study is the contribution to outcome made by the dietitian’s regular visits and the resulting encouragement and attention to detail in the intervention group.” This is an issue common to many trials i.e. if the intervention has several components that are absent from the control arm, which component had an impact on the outcome? Nevertheless, this is evidence to support intensive individualised input from dieticians. It will also be interesting to see the cost-effectiveness data that the authors suggest is coming soon.

QIP disappoints

The EPOCH study assessed the effect of a quality improvement intervention in emergency abdominal surgery units in 90 UK hospitals in a stepped wedge cluster randomised controlled trial. The primary endpoint was mortality 90 days after surgery. The QI intervention consisted of a 37-element care bundle and provision of training of local QI leads, education meetings, formation of “change teams”, educational resources and promotional materials. Unfortunately there was no difference in mortality between the intervention group and the group with no intervention. Length of stay and readmissions were also no different, although some measures of quality did seem better such as documentation of risk prior to surgery. The overall negative result is surprising, but could reflect the differing impact of the intervention between hospital sites because different hospitals tackled different issues with varying speed and success. A more consistent and narrower intervention might have been more effective although certainly less empowering for the local team if the approach is too top down.

Gallstones still galling

One might imagine persistent abdominal pain after cholecystectomy to represent a residual biliary problem. But perhaps some people’s pain was never due to gallstones, especially since gallstones are commonly present but do not cause symptoms. Given the high prevalence of post-cholecystectomy pain, this begged the question of whether to do as many cholecystectomies if many patients do not appear to benefit symptomatically. Van Dijk et al designed the SECURE randomised controlled trial to assess whether a restrictive strategy for offering cholecystectomy was noninferior to usual care (cholecystectomy offered at surgeon’s discretion). In the restrictive strategy there were five set criteria for severe attacks that had to be met before cholecystectomy was offered. Over 1000 patients from 24 centres across the Netherlands were included in the study. The primary endpoint was patient-reported pain at 12 months. It is difficult to fault the study design.

The results, though, may be hard to swallow. More patients were pain-free in the usual care group, i.e. the restrictive strategy was statistically not noninferior. The caveat is that protocol deviations in the restrictive strategy group were common i.e. patients received the surgery even when the criteria were not met, but of course the analysis that matters is the intention-to-treat. Disturbingly a large proportion of patients who had cholecystectomy remained in pain. The only promising finding from this trial was that the restrictive strategy resulted in a lower proportion of patients having cholecystectomy (68% compared to 75% in the usual care group, p=0.01).

Levetiracetam for paediatric convulsive status epilepticus

The EcLiPSE trial is a large randomised controlled trial comparing levetiracetam and the traditional phenytoin for second-line treatment (benzodiazepines being first line) for paediatric status epilepticus. The primary outcome was time from randomisation to cessation of convulsive status epilepticus. The median time in the phenytoin group was 45 minutes and 35 minutes in the levetiracetam group (p=0.2). The safety profile was similar for the two drugs.

JAMA Internal Medicine

Smoking and the multi-disadvantaged

Who are the smokers in the US? Hardly the most riveting of research questions, but this study took an interesting approach. Leventhal et al assessed the disadvantage characteristics of smokers, former smokers, and never smokers in a survey population of over 25s from 2008 and 2017. The disadvantages they chose to look at included unemployment in the past year, heavy drinking, poverty, not completing high school, disability and psychological distress. The higher the number of disadvantages that a person had, the more likely the person was to be a current smoker. For those with 0, 1 or 2 disadvantages the odds of being a current smoker decreased over the time period, but not in those with 3 or more disadvantages.


Canagliflozin success

SGLT2 inhibitors are good for the heart in diabetes. For more proof of the pudding, you can now look to the boldly named CREDENCE trial. This double-blind trial compared canagliflozin with placebo in patients with type 2 diabetes mellitus and chronic kidney disease with albuminuria. All patients were treated with renin-angiotensin system blockade. The trial was terminated early after a planned interim analysis because patients in the canagliflozin group were experiencing a much lower rate of the primary outcome of end-stage kidney disease, doubling of the serum creatinine or death from renal or cardiovascular causes.


Sweet sweet policy

The lords of the PRIMEtime model (a method of estimating the impact of population lifestyle changes on non-communicable disease) assessed the impact of the UK government’s sugar reduction programme. These authors used the National Diet and Nutrition Survey to determine diet. They then assumed diets would not change over time and that the government programme was successfully implemented and resulted in the planned sugar reduction. Then they estimated the resulting calorie reduction and weight change. From this they established there would be a huge reduction in obesity and obesity related disease. This is quite inspiring. The main limitation of this analysis (as with any modelling study) is the assumptions made. For example, the results rely on there being no changes in production formulations as a result of the programme and that eating patterns do not significantly change.

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

Competing interests: None declared.