Wednesday, October 14, 2015

A Costly, Complex Year: Examining Healthcare Cost And End of Life Care

By: Joann Schulte  DO, MPH; Editorial Board Member 

      Almost two-thirds of bankruptcy filings have a medical cause and 14.6% of those interviewed for bankruptcy in 2007 had anill child1with a complex medical condition. An estimated 2.8 to 3.3 million individuals with vast medical bills file for bankruptcy each year2.  If the percentage of filers with children has remained constant, it means between 408,800 and 481,800 people seeking bankruptcy protection each year have children with complex illnesses severe enough to prompt the filings.
      Work published this early relased recently in Pediatrics from  (doi: 10.1542/peds.2015-0260) will help provide needed information for children’s medical care and costs in the last year of life.  A group of researchers from Boston and Philadelphia used data from the forty-one free-standing children’s hospitals to describe illness and costs for such children.  The hospitals submitted data to the Pediatric Health Information System database, an administrative database that contains inpatient utilization and cost data. Children are assigned a unique identifier to track them across multiple admissions. The researchers used the data to characterize the admissions and costs incurred among a cohort of 1,252 children who were admitted and died in calendar year 2012.  Infants less than a year of age were excluded to ensure a complete one-year review.
      Those children had complex medical conditions in nine organ systems:  cardiovascular, congenital/genetic, gastrointestinal, hematologic/immunologic, malignancy, metabolic, neuromuscular, renal and respiratory. The authors converted charges to costs, using an existing ratio for each hospital and they adjusted for inflation by using the Consumer Price Index.
The authors found that the children in their last year of life had medians of two admissions, twenty-seven hospital days and $142,562 in hospital costs. Total hospital costs were $392 million, of which 58% occurred in the final admission for these children to the hospital($228 million).
      In a multivariate analysis, children diagnosed with hematological/immunologic conditions have the greatest impact on hospital resource use.  Compared to children with other complex conditions, children with hematologic/immunologic conditions spend 45 more days in the hospital and accumulated $326,844 more in hospital costs.  The vast majority of children (987, 79%) spent time in the intensive care unit and most were mechanically ventilated (946, 76 %) during their last admissions.
      These 1,252 children had an unhappy outcome, but information about their hospitalizations is an important step in painting a detailed picture about terminal pediatric illnesses.  Having such information may help economists and pediatric providers both provide needed care and figure out how to do so without bankrupting families.  


1. Himmelstein DU, Thorne D, Warren E, Woolhandler S.  Medical bankruptcy in the United States, 2007:  Results of a national study.  American Journal of Medicine  2009; 122 (8):741-746.

2. Himmelstein DU, Warren E, Thorne D, Woolhandler S.  Market Watch: Illness and injury as contributors to bankruptcy.  http://content.healthaffairs

Tuesday, October 13, 2015

Substance Use and Beliefs in Middle School Influence Risk of Driving Under the Influence: A Concerning Set of Longitudinal Survey Data

By: Lewis First, MD, MS; Editor-in-Chief     

          Risk-taking behaviors are certainly high on our anticipatory guidance radar when we see adolescents for health maintenance visits.  Even though we are concerned that a risk-taking behavior in early adolescence can portend other risk-taking behaviors in later adolescence—has it ever really been demonstrated?  Ewing et al. (doi: 10.1542/peds.2015-1143) have done just that using surveys of over 1100 teens given at ages 12 and 14 looking at alcohol and marijuana use as well as beliefs about their use and then compared these results to how often these same teens at age 16 drive under the influence (DUI) or ride with someone else who is drinking and driving (RWDD).  Even beginning at age 12, if a teen felt positively about marijuana use, let alone by age 14 was using it or drinking alcohol—there is a dramatic higher risk of DUI AND RWDD in these teens. 
Do you broach these risk-taking behaviors at age 12?  Do you wait until teens are older?  If the latter, you may be too late in offering the education these teens need to make better choices when confronted by peers with risk-taking opportunities for substance use and abuse.  How early do you begin to discuss marijuana and alcohol use with preteens and have you found talking to preteens is making a difference in a positive sense?  Share with us your thoughts on this interesting, but sad look at early adolescence through a response to this blog, an e-letter or posting a comment on our Facebook or Twitter sites.

Related Links 

Monday, October 12, 2015

An Ugly Reality: A Study of Developmental Behavior Of Children Born From Sexual Violence

By: Lydia Furman, MD,  Assistant Editor

      In a recently released study in Pediatrics, Dr. Rouhani et al. (doi: 10.1542/peds.2014-3373) presented us with a must read article about the burdens and experiences of women who are raising children conceived through sexual violence in the Democratic Republic of Congo (DRC) (ref here).  My first thought on reading this incredible article was that one does not need to go to the Democratic Republic of Congo (DRC) to meet women who are raising children conceived through sexual violence. Their conclusions are logical and intuitive, yet ground breaking, and I believe I can apply their “lessons” in my own practice. These authors interviewed over 700 women who are raising a child conceived through sexual assault. They used a specific method of locating hard-to-reach populations, called respondent driven sampling, in which ten initial identified women each recruited three other women, who then recruited three other women, and so on until the full population was identified.
      Please read the article to immerse yourself in the study, because likely you will find other parts of the work more meaningful than what captured me. What impressed me most was the extraordinary resilience of the women and of the maternal-child relationship.  Over one third of women reported that their community stigmatized their child, and over one third reported that their community stigmatized them personally. “Perceived acceptance of the child” by spouse, community and family seemed surprisingly low to me, given that the entire community was exposed to known, ongoing and multiple sexual assaults. Yet, despite all of this, 80.5% of women had “positive regard” (questions drawn from the Parental Stress Scale) for their child and 73.0% had “high attunement” (defined as “…awareness of, sensitivity to, and responsiveness to the child’s needs…”). So in the face of incomprehensible stress and challenge, including seeing their assailant and recalling the assault when looking at their child, the great majority of women are raising their children with engagement and kindness, as best one can understand. The authors’ multivariate analysis examines this question quantitatively and expands understanding further.
      While the upheaval in The DRC is notorious for the horror of widespread sexual violence, utilized as a weapon of war, the problems that women must face in raising a child who is the product of a rape respect no geographic boundaries. An issue that clinicians practicing in the US may have to recognize among women raising a child conceived through sexual violence is the complex racism of skin tone, which rears its ugly head when mother and child have obviously different skin colors or eye colors, prompting family or non-family members to ask about paternity. In the United States (US) it may be more possible for women to avoid stigma by non-acknowledgement of the sexual violence (e.g. by asking for non-inclusion of the information in the child’s medical record), but this does not permit future providers to treat the mother optimally (since she may have depression, anxiety, or other sequelae that are then not recognized), and this also makes support of any discussion she may want or need to have about conceiving in the setting of sexual violence   almost impossible.
       While Dr. Rouhani and colleagues studied a population in whom approximately 40% of women experienced sexual violence, it is sobering to realize that the Centers for Disease Control and Prevention (CDC) reports that “nearly one in five (18.3%) of women (in the US) … reported experiencing rape at some time in their lives” ( ). While a continent and world away for some, this is a reality we cannot ignore. Ultimately prevention is the best cure, but in the meantime, removing stigma is critical.  Work is underway to support de-stigmatization of sexual violence in the DRC, for example through the Harvard Humanitarian Initiative (, and a brief search reveals relatively limited initiatives in the US, most primarily focused on domestic violence (for example, Triumph, a network for survivors of domestic violence, We have work to do, and the superb research of Dr. Rouhani and colleagues in the DRC is a wakeup call to all of us.

Related Links

Friday, October 9, 2015

Hospital Variation on Care Utilization by Children with Medical Complexity: Does It Happen and What Do We Do about It?

By: Lewis First, MD, MS; Editor-in-Chief      

          Thanks to the information we can get from payer claim databases, we can learn a lot about variations in care delivery regarding different patient populations. One of those populations is the group of children with medical complexity who receive care daily at our children’s hospitals.  So how consistent is the care delivery across hospitals? Ralston et al. (doi: 10.1542/peds.2014-3920), in an article being released this month in our journal, performed a retrospective population-based observational cohort study that examined payer claims of all children from 1 month to 18 years with medical complexity in Maine, New Hampshire and Vermont.  Sadly there is more variation in care across the 6 hospitals studied in these states than one might suspect.  Where are these variations occurring?  Virtually everywhere--inpatient, outpatient, in intensive care units, and in the emergency department.  Even use of ancillary tests like radiology showed substantive variation.
So what does a study like this suggest?  Drs. Thomson and Shah offer a commentary on variability in health care utilization that accompanies this study and suggest how we can learn from the results being shared in this and other care variation studies that our journal and many others seem to publishing on a frequent basis.  We encourage you to read both the study and commentary and gain a better understanding of how this data might identify best practices for these patients that can then be shared and standardized across hospitals and across states. \
 If your state has an all-payer claims database, you might want to explore what that database is saying about your utilization rates and compare them to the data in this study. If the focus of care nowadays is on managing populations with high quality and lower cost, it is studies like this one that can set the stage to make that happen. 

Related Links