Orkaby AR et al. Association of statin use with all-cause and cardiovascular mortality in US veterans 75 years and older. JAMA 2020 Jul 7; retrospective cohort study of about 327,000 patients (age, ≥75; mean age, 81; mostly white men) without prior statin use, without a prior cardiovascular event; in total 326K vets that were 75yr old and older and had never had a cardiovascular event those started on statin therapy did better!!! all-cause mortality was 78.7 per 1000 person years in those started on a statin and 98.2 per 1000 person-years in those not a on a statin. BUT devel in the details-- they say they looked at 7.2 million vets- then once you took out the people that had prior statin exposure, and removed the people with missing data, and then excluded those with a prior cardiovascular event, and threw out those individuals who died in the first 150 days -- you then get the 326K. and of those about 57k got a statin prescription and 269K never got a statin prescription. BUT those individuals who started a statin were more like to have the diagnoses codes of hyperlipidemia, diabetes, and hypertension AND less likely to have dementia. so the patients that benefit from a statin are over the age of 75, have CV risk factors and dont have dementia.. well isnt it shocking that those individuals benefited from statin therapy! out of the initial 7.2 million people you cut it down to the 7% or 57k that got a statin and you say “look they did better” than the people that had fewer risk factors and were demented that didnt get a statin. OR the statin had nothing to do with it -- we will never know cause you can match up the confounding variables in retrospective data, no matter how hard to yu try!! NEVER. At this point statins are safe, we dont know if they work in the elderly because the data and studies often exclude them BUT we dont need any more observational studies, especially not ones that looks at 7.2 million charts then cuts it down to 326K. That is cherry picking the data and the cofounders will never be equal, at this point it is RCT or nothing, do the trial we all want or dont do the trial. Family med- jack of all and master of many—because you can do so many things- I am bias- Wallhed Finn S et al. Treatment of alcohol dependence in primary care compared with outpatient specialist treatment: Twelve-month follow-up of a randomized controlled trial, with trajectories of change. J Stud Alcohol Drugs 2020 May; moderate levels of alcohol dependence are more likely to seek care in primary care than a specialist randomized controlled noninferiority trial researchers in Sweden randomized 288 fulfilling ICD-10 criteria for alcohol dependence to comprehensive treatment by a team of experienced specialists at a single center or brief interventions provided by general practitioners who had received 8 hours of training. primary outcome was change in weekly alcohol consumption measured in grams of alcohol before inclusion compared with 12 months after start of treatment to give you and idea the averahe baseline alcohol consumption of these individuals was around 350G a week which equals about a 25 drinks a week the results were that no matter which group you were in – at 12 months you were drinking less – you wer down to about 12.5 drinks a week if you went to a specialist and about 13.5 drinks a week I guess this study fit my confirmation bias and specialist are not needed for anything ever- I am kidding of course BUt I am not kidding when I say I think screening for cancer is often a waste of time, at baseline people are healthy so it is very hard to screen for something when the person is healthy! conveying this message to patients can be very challenging. It is hard to explain risk and benefits. When talking to my patients I try to lay the numbers out – make like dr.(continued)