Zackary Berger’s weekly research reviews—14 January 2019

Zackary Berger reviews the latest research from the top medical journals


Medical marketing in the United States

US$30 billion dollars—that’s the total spent on medical marketing in the US in 2016, an increase from US$18 billion in 1997. Two Dartmouth researchersLisa Schwartz, whose recent premature death was widely mourned, and Stephen Woloshinquantified how the money spent on medical marketing has changed over those years, using a literature search and queries to industry, government, and NGOs. The most breakneck pace was set by direct to consumer pharmaceutical advertising, which increased from $2.1 billion (11.9%) of total spending in 1997 to $9.6 billion (32.0%) of total spending in 2016. But old reliable methods of marketing still make themselves felt: $979 million was spent in 2016 on direct payments to physicians. Laboratory testing is a widespread topic of advertising; 64% of direct to consumer advertising in 2016 was spent convincing consumers to undergo genetic testing, mostly for reasons having little to do with the promotion of health or the prevention of disease. Schwartz and Woloshin’s data gathering and searching provides important food for thought and guidance for policy.

Are there enough primary care clinicians in the USA?

How many clinicians do we need? The US Federal government assesses this on the basis of doctor:patient ratio, the poverty level in the geographic area in question, and the average distance to the nearest provider. According to estimates from the US Department of Health and Human Services in 2013, there aren’t enough primary care providers in the USA, and the shortage will get worse. What to do? A group of researchers in Rochester, New York, and San Francisco, used databases of physicians and nurse practitioners, comparing the trends in numbers and density of physicians and nurse practitioners between 2010 and 2016. They analyzed differences in these trends between rural and urban areas, as well as across levels of income.

The conclusions are clear. There are more physicians than nurse practitioners; however, the number of physicians is (broadly speaking) flat, while the number of nurse practitioners is increasing, especially in poorer and more rural regions. The analysis is incomplete as there is no all-encompassing registry of all providers, and cohort trends can be confounded. Since for most things physicians and nurse practitioners provide equivalent care, perhaps nurse practitioners really are the future saviour of access to primary healthcare in the US.

JAMA Internal Medicine

A home-based intervention to reduce disability

A disclosure: I work at the same institution as the authors of this study, and I have referred my patients to the intervention tested in this randomized control trial. The question at hand: can a home-based intervention, with attention to the personal goals of older people reduce disability? This study focused on older people who live at home (65 years and older) with low income. They had to be cognitively intact, not living in an apartment, and not undergoing cancer treatment or already working with physical or occupational therapists. They must have reported difficulty with either one activity of daily living or two instrumental activities of daily living.

First came multidisciplinary assessment by an occupational therapist and a registered nurse relating to function and safety. This assessment was then tailored to participants’ personal goals. Finally, implementation was carried out through home repairs, environmental modifications, and medical equipment. In the control group, a research assistant provided social home visits. The findings suggest that this affordable intervention (an average cost of about $2,000 per patient) was effective: self-reported disability was reduced at five months by 30% compared to the control group. But there are multiple limitations to the study: only about 25% of those screened for participation were eventually randomized in the trial. While the control was matched for time, the analysis was not blinded, and neither were the participants. As in any real world intervention, the question is what part of the multimodal package made the difference. Even further, in the practice setting where I see patients most days, the selection of patients is perhaps even stricter than in this trial. While this RCT can rightfully claim that it tested an intervention which worked in poor, older patients, as the project expands, sustainability and generalizability are questions that need to be addressed.

Opioids, pneumonia, and HIV: What causes the association?

Is the prescription of opioids associated with community-acquired pneumonia, and how does the association among people living with HIV compare to those without HIV? This study takes as its point of departure the fact that some opioids are more immunosuppressive than others, and that people living with HIV are more susceptible to a variety of infections. Using a cohort of patients from Veterans Administration hospitals and supplemental data from the US Center for Medicare and Medicaid Services, the researchers performed a nested case-control study: community-acquired pneumonia was the case, and opioids were the exposure. The study statistics control for relevant covariates and relevant sensitivity analyses were carried out. The hypothesized association between opioid prescriptions and community-acquired pneumonia checks out, and the association is stronger both for more immunosuppressive opioids (e.g., morphine compared to oxycodone) and for people living with HIV compared to those without HIV. But is this due to immunosuppression or some other association of opioids, such as pain? Chronic pain is often under recorded or not coded. Do we know whether chronic pain predisposes people to infection? While the association of opioids with adverse effects is important, potential mechanisms seem multiple and worth exploring.

Associations of intimate partner violence, sexual assault, and posttraumatic stress disorder with menopause symptoms

Sometimes the most important studies are surveys asking about common conditions which many people overlook. A group of researchers in Northern California did a cross-sectional survey of women aged 40-80 years old to find out how common are physical and emotional intimate partner violence, sexual assault, and post-traumatic stress disorder, and whether these are associated with symptoms of menopause. The lifetime prevalence of intimate partner violence (emotional 21%, physical 16%), sexual assault (19%), and current symptoms of post-traumatic stress disorder (23%) were unfortunate and unsurprising. The researchers found that these types of emotional trauma were associated with sleep difficulties, vasomotor symptoms, and vaginal symptoms accompanying menopause.

While no causal conclusions can be drawn, the authors discuss some implications which primary care providers must address: the importance of emotional trauma, across racial groups and in women without a high prevalence of depression and anxiety, and its connection to symptoms that are often potentiated by physical and emotional stress. The authors also suggest that trauma-informed care, a much-discussed topic whose literature is still nascent, might help address these associations.

Annals of Internal Medicine

Finding the balance between benefits and harms when using statins for primary prevention of cardiovascular disease

The evidence for the use of statins for the primary prevention of cardiovascular disease is not incontrovertible, yet rare is the patient or clinician who does not feel pressure to take or prescribe a statin. Much has been written about statins, but comparatively little space has been devoted to considering patient preferences. How can these be included when deciding whether to recommend statins? This study tries to calculate the balance of benefits and harms, and weigh these by patient preferences. Benefits were based on published systematic reviews. Patient preferences were estimated on the basis of another study (carried out by these same authors), in which participants in Ethiopia and Sweden were asked to rank preferences for various adverse health events. This methodology is well developed, but still depends on multiple assumptions, among them that real-life choice of treatment is done rationally, and that actually asking patients about their preferences at the time of decision, informing them about the uncertainty of the benefits involved, is somehow inferior to modeling such preferences on the basis of patients from two countries. Nevertheless, the authors used this model to find the level of cardiovascular risk, for various subgroups, at which the benefit/risk balance would tip towards statins. They found that such risk levels were higher than specified in current guidelines, i.e. they found that the benefits of statins tend to outweigh harms for patients at higher risk than currently recommended. This is interesting, but to my mind does not outweigh the immediate usefulness of asking the patient what they want to do.


Operative treatment versus nonoperative treatment of Achilles tendon ruptures

The Achilles tendon is often strained or sprained. But sometimes, with exertion there’s a tell-tale pop and severe shooting pain which herald an Achilles rupture. Then what? There have been a number of recent, dueling studies, some showing improvement through operative (surgical) repair, and othersnot so much. This is the perfect situation for a systematic review and, where possible, a meta-analysis. The authors searched the literature in April of last year, including 29 studies (going back to 1973). They asked whether operative or nonoperative correction (e.g., rehabilitation) was associated with lower rates of repeat rupture. They also included secondary outcomes such as return to work and sport. They performed a separate meta-analyses of randomized control trials and of observational studies, and comparing these estimates. It is not clear whether those performing the meta-analyses were blinded to the intervention of the studies. Operative repair reduced rates of repeat rupture by 1.6% (2.3% compared to 3.9%), but increased complications by 3.3% (4.9% compared to 1.6%). Whether there are patient subgroups more susceptible to repeat rupture, and whether patient inclusion in research design might have picked a different primary outcome, and included different studies, are questions for the future.

Zackary Berger is an associate professor at Johns Hopkins School of Medicine in the division of general internal medicine, and core faculty in the Johns Hopkins Berman Institute of Bioethics, both in Baltimore, Maryland. He is also a practising physician at a free health clinic for undocumented, predominantly Spanish-speaking, immigrants.

Competing interests: None declared