Roxanne Nelson, RN, BSN
December 28, 2018
According to a new observational study of more than 60,000 American adults, such a correlation exists.
A secondary analysis of the landmark Prostate, Lung, Colorectal and Ovarian Cancer Screening (PLCO) trial has found that all-cause mortality was higher in participants who did not adhere to recommendations for baseline cancer screening tests compared to those who did.
In fact, the authors observed that overall mortality (excluding deaths from cancers studied in the trial) was substantially higher among both patients who partially adhered to screening recommendations and those who did not at all adhere to the recommendations, as compared with persons who complied fully.
Specifically, during 10 years of follow-up, the hazard ratio of mortality, excluding deaths from cancers studied in the PLCO trial, and controlling for age, sex, and race/ethnicity, was 1.73 (95% confidence interval [CI], 1.60 - 1.89) for nonadherent compared with fully adherent participants and 1.36 (95% CI, 1.19 - 1.54) for partially adherent compared with fully adherent participants.
It is unclear how cancer screening can affect all-cause mortality, but an investigator speculated about the results.
"The most cogent explanation for these findings is that nonadherence to protocol screenings was a marker for a general behavioral profile of nonadherence to medical tests and treatments," said study coauthor Paul Pinsky, PhD, Division of Cancer Prevention, the National Cancer Institute, in comments to Medscape Medical News.
The increased risk is not related to the screening tests themselves, because deaths from the PLCO trial cancers were excluded in the all-cause mortality tally, he added.
The take-home message for clinicians, said Pinsky, "is that nonadherence with medical procedures is a significant risk factor for mortality, on par in terms of magnitude of excess risk with morbid obesity, and interventions to deal with this risk factor need to be developed and tested."
The new study was published online December 28 in JAMA Internal Medicine.
In an accompanying editorial, Deborah Grady, MD, MPH, and Monica Parks, MD, both of the University of California, San Francisco, point out that this was not a randomized trial that compared those who followed screening recommendations with those who did not. Rather, it was observational, and "like all observational studies, is susceptible to confounding.
"There is no way that nonadherence with cancer screening could cause increased mortality from a range of diseases not associated with screening," they write.
The editorialists agree with the investigators: it is most likely that nonadherence with recommended screening is a marker for behaviors that are associated with increased mortality.
"Previous studies have shown that patients who are adherent to recommended medications are more likely to seek out other preventive services such as screenings and vaccinations, while nonadherence has been associated with increased mortality," the editorialists write.
So do these results have any clinical relevance?
Yes and no, editorialist Grady told Medscape Medical News.
On one hand, she said, "the relevance is that the study demonstrates that in observational studies where people are not randomized to the groups being compared, there may be factors that are not measured or are difficult to measure that account for the association between the predictor — here, choosing to follow advice for screening — and the outcome — death, in this case.
"It's a warning to take observational findings with a grain of salt, especially in situations where the intervention requires people to do something or follow advice," she added.
Clinicians generally should not base clinical care on observational studies without confirmation from a randomized trial, Grady advised.
The PLCO trial was a large trial that investigated screening for four types of cancers. In the current study, Pinsky and coauthor Dudith Pierre-Victor, PhD, also from the National Cancer Institute, conducted a secondary analysis to assess mortality that was unrelated to the cancers being screened for in the PLCO trial.
"We use adherence to cancer screening here as an example because we have a large cohort, the PLCO trial cohort, followed for mortality for a long time and for whom we have information on such adherence," Pinsky said. "However, we believe we would see similar findings if adherence to other medical procedures was assessed."
A total of 64,567 (29,537 women and 35,030 men; mean age, 62.3 years) were eligible and were included in the current analysis. The majority of participants were adherent to the screening protocol (n = 55,065; 85.3%); 2548 (3.9%) were partially adherent; and 6954 (10.8%) were nonadherent. Those who were partially adherent completed a mean of 2.9 of four possible tests, and 74% only missed a sigmoidoscopy.
At 15 years' follow-up, 7966 deaths occurred (excluding those from the cancers being screened for) among patients who were fully adherent, 449 deaths among those who were partially adherent, and 1395 deaths among those who were nonadherent. This extrapolated to mortality rates (per 10,000 person-years) of 116.3 for patients who were fully adherent, 145.0 for those who were partially adherent, and 168.3 for those who were nonadherent.
After adjusting for medical risk factors for mortality and behavioral factors, the above-cited hazard ratio decreased to 1.46 (95% CI, 1.34 - 1.59) for nonadherent compared with fully adherent participants.
Pinsky noted that the PLCO cohort has been shown to display a "healthy volunteer effect." Inasmuch as the participants had volunteered to take part in a randomized trial, it is likely that they were more health conscious than the average US population.
"Still, 11% did not comply with any of the protocol screening tests at baseline, even though presumably healthy at the time," he noted. "The proportion with this 'nonadherence phenotype' in the general population therefore is probably greater than 11%. So, there is a substantial proportion of the population that has a health behavior phenotype that predisposes them to a very substantial increased risk of death over the next 10 years or so, on the order of 50% to 70% extra risk."
No outside funding for the study was reported. The authors and editorialists have disclosed no relevant financial relationships.
Health Solutions From Our Sponsors
SOURCE: Medscape, December 28, 2018. JAMA Intern Med. Published online December 28, 2018.