Paying Family Medical Caregivers for Children's Home Healthcare in Colorado: A Working Medicaid Model.

OBJECTIVE: To compare the characteristics and healthcare use of children with medical complexity who receive paid certified nursing assistant (CNA) care by a family member (family CNA) and by a traditional nonfamily member (nonfamily CNA). STUDY DESIGN: This was retrospective cohort study of children who received CNA care through Colorado's Medicaid paid family caregiving program between 2017 and 2019 by a home healthcare agency. We compared patient characteristics between the family CNA and nonfamily CNA groups. A multivariable Poisson regression model was used to compare hospitalization rates (days in the hospital per year), adjusting for patient age patient sex, nursing care, and complex chronic condition. RESULTS: Of 861 patients, 79% (n = 680) received family CNA care and 21% (n = 181) received nonfamily CNA care. Patient demographics and hospitalization did not differ between the groups, although patients who had family CNAs were less likely to receive additional nursing-level care (42% vs 60%, P < .01). Family and nonfamily CNA caregivers had similar characteristics, except that family CNA caregivers had substantially better 3-year retention (82% vs 9%, P < .01) despite lower average hourly pay ($14.60 vs $17.60 per hour, P < .01). Hospitalizations were rare (<10% of patients). In the adjusted model, patients who received family CNA care experienced 1 more hospitalized day per year, compared with patients who received nonfamily CNA care (P < .001). CONCLUSIONS: Paid family caregivers provided CAN-level care to children with medical complexity with a greater employee retention compared with nonfamily CNA caregivers, with marginally different hospitalization rates using a family-centered approach. This model may help address workforce shortages while also providing income to family caregivers.

Quality of Care for Youth Hospitalized for Suicidal Ideation and Self-Harm.

OBJECTIVE: To examine performance on quality measures for pediatric inpatient suicidal ideation/self-harm care, and whether performance is associated with reutilization. METHODS: Retrospective observational 8 hospital study of patients [N = 1090] aged 5 to 17 years hospitalized for suicidal ideation/self-harm between 9/1/14 and 8/31/16. Two medical records-based quality measures assessing suicidal ideation/self-harm care were evaluated, one on counseling caregivers regarding restricting access to lethal means and the other on communication between inpatient and outpatient providers regarding the follow-up plan. Multivariable logistic regression assessed associations between quality measure scores and 1) hospital site, 2) patient demographics, and 3) 30-day emergency department return visits and inpatient readmissions. RESULTS: Medical record documentation revealed that, depending on hospital site, 17% to 98% of caregivers received lethal means restriction counseling (mean 70%); inpatient-to-outpatient provider communication was documented in 0% to 51% of cases (mean 16%). The odds of documenting receipt of lethal means restriction counseling was higher for caregivers of female patients compared to caregivers of male patients (adjusted odds ratio [aOR] 1.51, 95% confidence interval [CI], 1.07-2.14). The odds of documenting inpatient-to-outpatient provider follow-up plan communication was lower for Black patients compared to White patients (aOR 0.45, 95% CI, 0.24-0.84). All-cause 30-day readmission was lower for patients with documented caregiver receipt of lethal means restriction counseling (aOR 0.48, 95% CI, 0.28-0.83). CONCLUSIONS: This study revealed disparities and deficits in the quality of care received by youth with suicidal ideation/self-harm. Providing caregivers lethal means restriction counseling prior to discharge may help to prevent readmission.

Prevalence of Social Risks on Inpatient Screening and Their Impact on Pediatric Care Use.

OBJECTIVES: Screening for social determinants of health in the inpatient setting is uncommon. However, social risk factors documented in billing and electronic medical record data are associated with increased pediatric care use. We sought to describe (1) the epidemiology of social risks and referral acceptance and (2) association between social risks identified through routine inpatient screening and care use. METHODS: Parents of children ages 0 to 18 admitted to a general pediatric floor at an academic children's hospital completed a psychosocial screening survey from October 2017 to June 2019. The survey covered the following domains: finances, housing, food security, medications, and benefits. Patient characteristics and care use outcomes were abstracted from the electronic medical record and compared by using Pearson's χ2 or the Wilcoxon rank test and logistic regression analyses. RESULTS: Of 374 screened families, 141 (38%) had a positive screen result, of whom 78 (55%) reported >1 need and 64 (45%) accepted a community resource. In bivariate analyses, patients with a positive screen result had higher 30-day readmission (10% vs 5%; P = .05), lower median household income ($62 321 vs $71 460; P < .01), lower parental education (P < .01), public insurance (57% vs 43%; P < .01), lived in a 1-parent household (30 vs 12%; P < .01), and had a complex chronic condition (35% vs 23%; P = .01) compared with those with a negative screen result. There was no difference in care reuse by screening status in adjusted analyses. CONCLUSIONS: Social risks are common in the pediatric inpatient setting. Children with medical complexity offer a good target for initial screening efforts.

How Measurement Variability Affects Reporting of a Single Readmission Metric.

Readmissions are an important quality measure for public reporting, payment, and collaborative research. Lack of measure standardization may lead to inconsistent reporting of outcomes across study sites. In this study, we examined the impact of measurement variability on reporting of a single readmission metric, 30-day all-condition readmission rates (ARRs). We conducted a secondary database analysis of 2006-2008 Medicaid Analytic eXtract data merged from four states of children younger than 21 years. We calculated 30-day ARRs for this cohort using three previously described models varying in their inclusions and exclusions of index hospitalizations and readmissions. The 30-day ARR was highest for the model allowing each readmission to serve as an index admission for subsequent readmissions (ARR: 7%); intermediate for the model allowing one index admission and more than one readmissions in each 30-day period (ARR: 6.2%); and lowest for the model allowing only one readmission in each 30-day period (ARR: 5.6%). Similar variation was seen when stratifying patients by individual diagnostic groups. In conclusion, measurement variability impacts reported outcomes of a single readmission metric. To improve the value of readmission as a quality metric, stakeholders engaged in multisite quality improvement or research should ensure that definitions are standardized across sites.

Potential Impact of Initial Clinical Data on Adjustment of Pediatric Readmission Rates.

OBJECTIVE: Comparison of readmission rates requires adjustment for case-mix (ie, differences in patient populations), but previously only claims data were available for this purpose. We examined whether incorporation of relatively readily available clinical data improves prediction of pediatric readmissions and thus might enhance case-mix adjustment. METHODS: We examined 30-day readmissions using claims and electronic health record data for patients ≤18 years and 29 days of age who were admitted to 3 children's hospitals from February 2011 to February 2014. Using the Pediatric All-Condition Readmission Measure and starting with a model including age, gender, chronic conditions, and primary diagnosis, we examined whether the addition of initial vital sign and laboratory data improved model performance. We employed machine learning to evaluate the same variables, using the L2-regularized logistic regression with cost-sensitive learning and convolutional neural network. RESULTS: Controlling for the core model variables, low red blood cell count and mean corpuscular hemoglobin concentration and high red cell distribution width were associated with greater readmission risk, as were certain interactions between laboratory and chronic condition variables. However, the C-statistic (0.722 vs 0.713) and McFadden's pseudo R2 (0.085 vs 0.076) for this and the core model were similar, suggesting minimal improvement in performance. In machine learning analyses, the F-measure (harmonic mean of sensitivity and positive predictive value) was similar for the best-performing model (containing all variables) and core model (0.250 vs 0.243). CONCLUSIONS: Readily available clinical variables do not meaningfully improve the prediction of pediatric readmissions and would be unlikely to enhance case-mix adjustment unless their distributions varied widely across hospitals.

Hospital Utilization Among Children With the Highest Annual Inpatient Cost.

BACKGROUND AND OBJECTIVES: Children who experience high health care costs are increasingly enrolled in clinical initiatives to improve their health and contain costs. Hospitalization is a significant cost driver. We describe hospitalization trends for children with highest annual inpatient cost (CHIC) and identify characteristics associated with persistently high inpatient costs in subsequent years. METHODS: Retrospective study of 265 869 children age 2 to 15 years with ≥1 admission in 2010 to 39 children's hospitals in the Pediatric Health Information System. CHIC were defined as the top 10% of total inpatient costs in 2010 (n = 26 574). Multivariate regression and regression tree modeling were used to distinguish individual characteristics and interactions of characteristics, respectively, associated with persistently high inpatient costs (≥80th percentile in 2011 and/or 2012). RESULTS: The top 10% most expensive children (CHIC) constituted 56.9% ($2.4 billion) of total inpatient costs in 2010. Fifty-eight percent (n = 15 391) of CHIC had no inpatient costs in 2011 to 2012, and 27.0% (n = 7180) experienced persistently high inpatient cost. Respiratory chronic conditions (odds ratio [OR] = 3.0; 95% confidence interval [CI], 2.5-3.5), absence of surgery in 2010 (OR = 2.0; 95% CI, 1.8-2.1), and technological assistance (OR = 1.6; 95% CI, 1.5-1.7) were associated with persistently high inpatient cost. In regression tree modeling, the greatest likelihood of persistence (65.3%) was observed in CHIC with ≥3 hospitalizations in 2010 and a chronic respiratory condition. CONCLUSIONS: Most children with high children's hospital inpatient costs in 1 year do not experience hospitalization in subsequent years. Interactions of hospital use and clinical characteristics may be helpful to determine which children will continue to experience high inpatient costs over time.

Outpatient follow-up visits and readmission in medically complex children enrolled in Medicaid.

OBJECTIVE: To examine the association between postdischarge outpatient follow-up and 30-day readmissions in Medicaid enrolled children with complex, chronic conditions. STUDY DESIGN: This was a retrospective cohort analysis of Colorado Medicaid recipients with complex, chronic conditions who were discharged from the hospital between 2006 and 2008. The primary outcome was readmission between 4 and 30 days after index hospital discharge. Using multivariable logistic regression, we examined the association between early postdischarge outpatient visits (≤ 3 days postdischarge) and readmission. We secondarily analyzed the relationship between any outpatient visit from 4 to 29 days of index discharge and readmission. RESULTS: For the 2415 patients with complex, chronic conditions included in the analysis, the 4- to 30-day readmission rate was 6.3%. The odds of readmission was significantly greater for patients with ≥ 1 outpatient visit ≤ 3 days after discharge compared with patients without a visit ≤ 3 days after discharge (aOR 1.7 [1.1-2.4]). The odds of readmission were significantly lower for patients with ≥ 1 outpatient visit from 4 to 29 days after discharge compared with patients without such visits (aOR 0.5 [0.3-0.7]). Other factors associated with readmission included index hospital length of stay and number of complex, chronic conditions. CONCLUSIONS: In medically complex children, there is a positive association between early postdischarge outpatient follow-up and readmission. There is an inverse association between later postdischarge outpatient follow-up and readmission. Outpatient follow-up occurring within 4-29 days after discharge may help to prevent 30-day readmissions. Additional research is needed to inform guidelines regarding longer term postdischarge outpatient follow-up in these children.