NEJM Perspective praises work of Mini-Sentinel Program

Despite the expanding numbers of agencies with dense electronic medical data, the research results from claims data and EHRs are seldom useful, partly because these medical big data are not of research quality but rather are the electronic side effects of the functions of billing and clinical care.
 
The NEJM Perspective article highlights the Mini-Sentinel project as one system that can contribute meaningfully to the health of the public.  A pilot project of the Sentinel Initiative of the Food and Drug Administration (FDA), the Mini-Sentinel, has created a nationwide system that uses electronic data to evaluate the safety of marketed drugs, devices, and biologics.  Mini-Sentinel's coordinating center is located at the HPHC Institute and led by DPM Chair Richard Platt, MD, MS.
 
To be eligible to become a Mini-Sentinel data partner, health care systems are required to have complete longitudinal claims data, including computerized pharmacy data, for defined populations during known periods, as well as the ability to retrieve full-text medical records. Unlike fee-for-service providers, these closed systems of care are unlikely to miss key exposures or events in longitudinal studies.
 
Local data sets from data partners, formatted according to the common data model, constitute the Mini-Sentinel distributed database which, as of September 2013, included about 150 million covered lives, 360 million person-years of observation, 4 billion pharmaceutical dispensings, and 4.1 billion encounters. Each month adds another 45 million dispensing and 51 million encounters.
 
The NEJM piece discusses how actively assessing the risk–benefit profile of drugs in the post-market setting is a new activity for the FDA, and the agency may be considering how to integrate the Mini-Sentinel into the regulatory process.
 
In all, the Mini-Sentinel, which costs about 6 cents per capita per year, protects privacy, maintains transparency, and provides an essential public health service. According to the Perspective authors, the current configuration — the data model, the methods development, and the investigative team-- represents an impressive achievement that deserves ongoing support when the Sentinel Initiative comes up for renewal this fall.

Digital Mammography Modestly Improves Health But Increases Costs

In a comparative modeling study, conducted by the Harvard Pilgrim Health Care Institute and National Cancer Institute -sponsored consortia, researchers used five independent breast cancer simulation models, developed through the Cancer Intervention and Surveillance Modeling Network, to examine the health benefits and harms as well as economic costs of screening strategies using digital mammography. In addition to examining the impact of using digital instead of film for routine screening, researchers also examined digital screening strategies starting at age 40 or 50 applied annually, biennially, or based on mammographic breast density. Mammography performance and information on breast density were based on data from the Breast Cancer Surveillance Consortium.  Model outcomes included mortality reductions, life expectancy, quality-adjusted life expectancy and costs.  Incremental cost-effectiveness ratios comparing strategies were computed.
 
Results demonstrated that routine use of digital mammography modestly improved health relative to film mammography, but resulted in increased overall cost largely due to higher reimbursement for digital and more false positive workups.
 
According to researcher and first author Natasha Stout, PhD  of the Department of Population Medicine at Harvard Pilgrim Health Care Institute and Harvard Medical School, the impact of these unintended consequences on population health and medical costs should be evaluated before widespread adoption of any new technology.
 
The researchers also found that a policy to extend biennial screening with digital mammography to women 40-49 years could be a reasonable value especially for women who are not adversely affected by the prospect of a false positive test.  Across the 5 models, extending biennial digital screening to women ages 40-49 relative to screening women 50-74 would lead to a median gain of 5 days in life expectancy, avert 1.1 additional deaths per 1000 women screened and increase costs by $0.69 million. However, when quality of life effects such as anxiety from screening and possible false positive results or invasive diagnostic workups are considered for women in their 40s, the small gain in health is reduced and this policy is less attractive from a cost-effectiveness perspective.  Screening annually from ages 40-74 resulted in the greatest health gains but also markedly higher costs and false positives than biennially screening.

New Approach to Understanding Determinants of Cesarean Delivery

DPM Research Fellow, Lauren Wisk, PhD  and colleagues designed a study that takes a life course approach to understanding the factors contributing to delivery methods in the US. Three delivery methods were examined --vaginal delivery, medically-indicated C-section, and non-medically indicated C-section – using data from a large, national, population-based survey of over 9,000 women. 
 
The life course model suggests that perinatal health may be influenced by numerous events across the lifespan, as well as by intergenerational effects. Thus, using multinomial logistic regression, the researchers examined the role of socio-demographics, health, healthcare, stressful life events, pregnancy complications, and history of C-section on the odds of medically- and non-medically indicated C-sections, compared to vaginal delivery.
 
Results indicated that prior C-section was the strongest predictor of both medically- and non-medically indicated C-sections.  As such, clinicians need to discuss with their patients possible health problems associated with multiple C-sections and should follow guidelines of the American College of Obstetricians and Gynecologists.
 
But there were marked differences in the risk factors for indicated and non-indicated C-sections.  Compared to women delivering vaginally, those who had unnecessary C-sections were more likely to have had a prior C-section, one or more prior live births, pre-pregnancy obesity, or were 35+ years old.  To reduce such inappropriate utilization, best practice guidelines should be discussed with doctors and patients and provider incentives should be examined.
 
Women whose C-sections were medically-indicated were more likely to have had any pregnancy complication, any chronic condition, given birth to multiples, no prior live births, and were the least likely to live in the West compared to women who delivered vaginally. Here, clinical or programmatic activities may need to be implemented prior to conception.
 
Surprisingly, stressful life events during pregnancy, health insurance during pregnancy, the time prenatal care was started, maternal race/ethnicity, marital status, SES, and urbanicity were not associated with delivery method.

Mandates for Sepsis Care: Proceed with Caution

Many studies report that the incidence of sepsis, the state of body-wide inflammation that occurs with severe infection, has increased dramatically over time.  It is a leading cause of death and was the single most costly condition treated in hospitals in 2011. 
 
Standardized protocols aimed at improving sepsis identification and treatment have become common in hospitals around the country. Policy makers, starting with New York State, have begun mandating hospital adoption and adherence to these protocols, and now the Centers for Medicare and Medicaid Services is considering adopting a metric for protocol adherence that was recently ratified by the National Quality Forum.
 
But according to Chanu Rhee, MD, Michael Klompas, MD, MPH and colleague, while the increasing attention to sepsis care is a positive step, mandated policies must be tempered by two limitations: (1) that we still need to learn the best way to diagnose and treat this condition and (2) that we don’t yet have reliable surveillance tools for measuring sepsis incidence.  Without accurate diagnostic tools, sepsis mandates can be risky because the nonspecific symptoms of sepsis may result in indiscriminate prescribing of antibiotics, overuse of invasive catheters and fluids, and unnecessary utilization of scarce ICU resources.  Without accurate surveillance methods, benchmarking hospitals’ performance on sepsis care may be too easily influenced by variability and changing practices in how sepsis is diagnosed and reported.
 
To read more details in the full NEJM Perspective, click here.

A Health Insurance Fine on Obesity: Ineffective and Discriminatory

In a post on The Health Care Blog, Stephen Soumerai, ScD, Professor of Population Medicine and Director of the Drug Policy Research Group discusses how this year,   under the Affordable Care Act, companies can assess 30% surcharges to employees with medical conditions such as obesity, high blood pressure, or high cholesterol.  The supposed rationale is to use financial penalties as incentives to change unhealthy behaviors, thus reducing health care costs, but no evidence exists to support such benefits and the charges may do more harm than good.

Dr. Soumerai reports that the obesity epidemic is extremely complicated and typical “wellness” programs do not achieve meaningful weight reduction. For example, there is a predisposition to obesity, making it much easier for some to gain weight, as well as childhood environmental factors that lead to obesity and diabetes -- conditions that are extremely difficult to reverse.

According to the author, fining overweight people is immoral and disrespectful, and may even be a violation of federal discrimination laws.  Rather than fining individuals, we need to focus on population-based prevention efforts, such as better access to healthy foods, increased school physical education budgets, and higher taxes on high calorie beverages.

 

 

Alternatives to Hospitalization for Frail Elders

According to Dr. Muriel Gillick, a physician who specializes in the care of elderly patients (geriatrics) and those with advanced illness (palliative care),trying to prevent geriatric medical problems that often lead to hospitalization – such as falls or episodes of heart failure-- is extremely difficult. But the hazards of hospitalization for frail patients can include adverse drug reactions, falls, delirium, and incontinence.
 
Dr. Gillick believes that when patients learn the risks of hospitalization and comprehend their frail status, they might decide to forgo hospital treatment if they have a reasonable alternative.  She describes some excellent options, including hospice and home hospital programs. She also touches upon the barriers to implementing hospital alternatives which include a Medicare minimum requirement for hospital stays; the convenience to clinicians of hospitalizing rather than arranging care with other facilities or home care givers; and patient wishes to avoid burdening their families with their care.
 
Dr. Gillick concludes that frail elders cannot be expected to avoid hospitals entirely, and that hospitalization might be the most appropriate course of treatment for some patients. But we need to minimize the need for hospitalization both with adequate transitional care, effective management of geriatric medical issues, and an integrated, coordinated system of care and also with viable alternatives to hospital treatment.
 
To view the Annals of Internal Medicine article click here.

 

Policy Changes Needed in Comparative Effectiveness Research

More than ever, our health care system can learn much about prevention, diagnosis, and treatment of illness due to an increased interest in comparative effectiveness studies and practice-based research, and the availability of data from large health information systems.  But, in the opinion of DPM’s Dr. Richard Platt and his co-authors, such approaches are inhibited by assuming that evaluation automatically creates higher risks to patients than usual care.
 
Presented in the article are a number of recommendations that could be considered within existing regulations:
 
Government agencies dealing with human research protection should provide more detailed information about permitted activities and precedents to assist IRBs in their decision-making. IRBs tend to err on the side of over-restriction because of their own uncertainty about the law.
 
Certain health care operations should be redefined, so that routine and continuous evaluation of health care processes are not classified as “research” and thus subjected to strict research regulations.
 
The risk-based approach to evaluating research studies needs to be modified.  Potential risks and burdens to patients in a study to better learn about the efficacy of a treatment need to be compared to the risks of care to patients without the benefit of such an evaluation.
 
The authors point out that systematically assessing care can result in significant opportunities to improve health outcomes and that research policies should enable, rather than inhibit, this potential.
 
To read the Viewpoint article, click here.

PCORnet

Summary

PCORnet will be a secure, highly representative, national network for conducting clinical outcomes research that improves the speed, efficiency, and use of patient-centered CER.

Learn More Link

www.pcornet.org

No Clinical Effects Seen with Changes in CABG Payment Policy

The Centers for Medicare and Medicaid Services (CMS) implemented policy changes in 2008 to eliminate additional Medicare coverage of mediastinitis following coronary artery bypass graft (CABG) surgery. To evaluate the impact of this policy on mediastinitis rates, Grace Lee, MD, MPH and colleagues from the Harvard Pilgrim Health Care Institute, as well as researchers from other institutions, conducted a study using Medicare claims and National Healthcare Safety Network (NHSN) prospective surveillance data.
 
The team compared mediastinitis rates (adjusted for secular trends) before and after the policy.  Billing rates were obtained from Medicare inpatient claims following over 600,000 CABG surgeries in more than 1200 hospitals in the United States. Prospective surveillance rates came from ~ 150 NHSN hospitals in 29 states performing close to 95,000 CABG procedures.
 
Results demonstrated an immediate decrease in coding for index admission mediastinitis when payment policy changes were implemented and a decreasing trend in coding for index admission mediastinitis in the post- versus pre-intervention period.
 
Interestingly, the new policy did not appear to affect infection rates as measured using NHSN data. The CMS policy of withholding additional Medicare payment for mediastinitis on the basis of claims-based evidence of infection resulted in changes in coding for infections, but did not change actual infection rates during the initial 2 years following policy implementation.
 
Click this link to read the abstract in Infection Control and Hospital Epidemiology.

Health Care and Population Health

In a recent JAMA Viewpoint, DPM faculty members Emma Eggleston, MD, MPH and Jonathan Finkelstein, MD, MPH discuss how health systems can contribute population health by taking more responsibility for the health of their patient populations, partnering with public health and other entities that influence health, and responding to societal pressure for equity and value.  The Viewpoint also addresses the challenges health systems face in achieving these goals.
 
To read the Viewpoint article, click here.
Subscribe to