Profit Status and Delivery of Hospice Care for Infants: The Mediating Role of Pediatric Knowledge

Introduction

Infant mortality is a persistent problem in the United States. ¹ Infants under 1 year of age represent approximately 55% of all pediatric deaths, a statistic relatively unchanged in 20 years.^2,3 Yet less than 20% of pediatric hospice admissions are infants. ⁴ Instead, many infants spend their short lifespan in the neonatal intensive care unit where they receive aggressive and invasive treatments.^5,6 This scenario is often contrary to the wishes of the family. ⁷

Previous literature suggests that the profit status of a hospice (i.e., nonprofit or for-profit) may play a role in the delivery of hospice care for infants. Nonprofit hospices may more readily take on patients with complex needs, such as patients with noncancer diagnoses, higher skill needs, and shorter length of stay.^8,9 Nonprofits also provide a wider range of hospice services,^10,11 and are more likely to provide care for children. ¹² Nonprofits may, therefore, also be more likely to deliver care for infants, who often have high skill needs, short lengths of stay, and noncancer diagnoses at the end of life. ¹³

The delivery of care to infants, however, may not be explained by profit status alone. Hospice characteristics such as possessing pediatric knowledge of how to care for children may mediate the delivery of care for infants. Knowledge is often formalized in an organization through rules/directives and routines. For example, the integration of rules and directives connected with a formal pediatric hospice program and the routines reinforced through the experience of providing pediatric care for any child may be a necessary component in the capability of a nonprofit or for-profit hospice to deliver infant care.^14,15

In light of the Concurrent Care for Children provision (Section 2302) of the Patient Protection and Affordable Care Act of 2010, a recent federal policy change impacting the delivery of hospice care for children that includes infants, it is important to understand the underlying mechanisms that influence the delivery of hospice care for infants among nonprofit and for-profit hospices. Exploring the role of pediatric knowledge may assist hospice clinicians and administrators in better targeting interventions designed to improve the quality of end-of-life care for infants. The purpose of our study was to examine the influence of profit status on delivery of hospice care for infants, and to explore the mediating role of pediatric knowledge on the relationship between profit status and the delivery of infant care.

Model and Hypotheses

Figure 1 represents the general conceptual model and hypothesized relationships among the study variables. We drew from organizational theory to evaluate the mediating role of pediatric knowledge on the relationship between profit status and delivery of infant care. Knowledge-based theory emphasizes that organizations develop capabilities through specialized knowledge found in the organizations' rules and directives and the routines of their employees. ¹⁶ In the case of hospice, the capability to deliver infant care may be facilitated by the rules and directives of a formal pediatric hospice program and the routines of pediatric-experienced staff. Thus, our conceptual model posits that profit status is directly related to the delivery of hospice care for infants, as well as indirectly related through pediatric knowledge. The proposed hypotheses are as follows:

OPEN IN VIEWER

FIG. 1.

Proposed model of the delivery of hospice care for infants.

Methods

Our study was a cross-sectional secondary analysis. We used the 2008 California Office of Statewide Health Planning and Development (CA OSHPD) State Utilization Data File of Home Health Agencies and Hospice Facilities to compile our hospice sample. The CA OSHPH data files provide a comprehensive description of the infrastructure and practices of hospices in California. Participation is mandatory for licensed hospice and home health agencies, and the survey is administered annually by the CA OSHPD. Data quality is assessed through an initial automated review process and then a more rigorous review of potential errors and inconsistencies by the CA OSHPD staff. ¹⁷ We included all hospices with active hospice program licensure and excluded hospices with duplicate entries, missing data, or both. The sample consisted of 254 hospices in California.

For this study, the outcome variable was the delivery of hospice care for infants. The 2008 CA OSHPD dataset had 12 age groups spanning from 0 to greater than 91 years of age. Hospice administrators were asked on the survey whether or not they admitted an infant ≤1 year of age to their care during a 12-month period. Hospices that responded “yes” were designated as hospices that delivered hospice care for infants. The primary independent variable was profit status. It was operationalized as whether a hospice reported its profit status as nonprofit or for-profit.

To explore the mediating effect of pediatric knowledge, two variables were evaluated. Hospices were asked on the survey whether or not they had a specialized pediatric hospice program. Those that answered “yes” were designated as having a pediatric program. Hospices were also asked to report the annual average number of children under 21 years of age cared for by the hospice. This figure served as the measure of pediatric experience.

Other organizational characteristics (i.e., facility size, service area, affiliation, organization age, competition) were included as covariates in our model. The facility size variable was operationalized as small size if a hospice had an average daily census ≤100 patients per day and large if they had an average daily census of >100 patients per day. Service area was categorized by whether hospices delivered care in urban or rural locations. Affiliation was measured categorically as whether hospices were freestanding or not freestanding (i.e., hospital based, home-health based, long-term care based). The total number of years a hospice was licensed was the measure for organizational age. Competition was measured with the Herfindahl-Hirshman Index (HHI). The measure was reversed (i.e., 1-HHI) to reflect the shift from monopoly to competition in the market of the hospice. A rating of 0 indicates a monopoly and a rating of 1.0 indicates extreme competition.

Descriptive statistics for all study variables were calculated. To examine the extent to which the relationship between profit status and the delivery of hospice care for infants was mediated by pediatric knowledge, a mediation analysis following the guidelines of Baron and Kenny was conducted. ¹⁸ First, the outcome variable (delivery of hospice care for infants) was regressed on the independent variable (profit status). Next, the relationships between profit status and each of the pediatric knowledge mediators (pediatric program and pediatric experience) were examined. Afterward, a regression model was used to estimate the simultaneous impact of profit status and pediatric knowledge on the delivery of hospice care for infants. Finally, the Sobel test was used to evaluate the significance of the indirect association through pediatric knowledge of profit status with the delivery of hospice care for infants. This study used both linear and nonlinear regression models depending on the distribution of the mediator and outcome variables. Results are reported as unstandardized coefficients for ordinary least squares and logit estimation methods. Analyses were performed using Stata version 11.0 (StataCorp LP, College Station, TX).

Results

Characteristics of the hospices studied are shown in Table 1. Of the hospices in the sample, 13% delivered care for infants. Less than half of the hospices were nonprofit (33%). Although few hospices had a specialized pediatric program (8%), hospices generally had experience with more than one child. The majority of hospices were small (83%), provided services in urban communities (60%), and operated as freestanding hospices (71%). On average, hospices were 9 years old with a range of 1 to 42 years. Hospices also operated in areas of relatively strong (0.77) competition.

The results of the mediation analysis are presented in Table 2. Compared to for-profit hospices, nonprofit hospices were more likely to delivery care for infants (B=0.67, p<0.05), have a specialized pediatric program (B=0.85, p<0.05), and have more experience caring for pediatric patients (B=1.49, p<0.05). Thus, our first hypothesis was supported.

Regarding the mediator variables, hospices with a pediatric program were no more likely to deliver care to infants than hospices without a pediatric program (B=0.82, p>0.05). This analysis suggested that a pediatric program did not mediate the relationship between profit status and delivery of care, and our second hypothesis was not supported.

Experience with greater numbers of any pediatric patients, however, was associated with the delivery of hospice care for infants (B=0.86, p>0.05). Additionally, when both profit status and pediatric experience were considered jointly, the likelihood of hospices delivering care for infants was reduced, indicating partial mediation. Using the Sobel test, pediatric experience was a partial mediator of the relationship between profit status and the delivery of hospice care for infants (Sobel=2.86, p<0.05), accounting for a relatively large portion of the total effect (86%). Therefore, the third hypothesis was supported.

Discussion

Our study investigated the influence of pediatric knowledge on the relationship between profit status and the delivery of infant hospice care. Although evidence suggests that hospice is underutilized for infants at the end of life, we know little about the factors that drive this process. We used a knowledge-based organizational theory to conceptualize and differentiate between hospice program rules and directives, operationalized as a formal pediatric program, and their routines, operationalized as pediatric experience. We found that nonprofit hospices were more likely to deliver hospice care for infants. Several studies have suggested that differences in the mission of nonprofit and for-profit hospices may influence the delivery of hospice care to high cost patients.^8–11 Nonprofit hospices may generally be more responsive to the unique needs of a local community, whereas for-profit hospices may tend to focus on core income-producing services. ¹⁹ In addition, nonprofit hospices are typically reliant on their communities for volunteers and funding and do not have stockholders to satisfy or profits to generate. ¹⁹ Thus, they can more easily play a critical role in providing a community benefit than for-profit hospices. This justifies the tax benefits they enjoy given their provision of care to underserved and vulnerable populations, including infants.

Surprisingly, we found that hospices with pediatric programs were no more likely to enroll infants than hospices without pediatric programs. However, experience with pediatric patients appeared to be instrumental in the delivery of hospice care for infants. Although our study did not allow us to understand the reasons for these findings, it is possible that the rules and directives of pediatric programs are developed with older children in mind. Pediatric programs are often structured to provide psychosocial and spiritual counseling, pet visitation, massage therapy, and school resources,^14,20 features that are suitable for school-age children and adolescents. In contrast, families with infants may need highly specialized services, such as infant bereavement and memory-making activities, that are common in perinatal hospice programs but may not be present in current pediatric programs for children.^21,22 It is possible that pediatric experience, on the other hand, may provide increased comfort in working with children of all ages. ²³ Future research might explore characteristics of pediatric programs to better understand their components' age group foci, as well as the aspects of pediatric experience that facilitate care for infants.

This study has several limitations. First, it was cross-sectional rather than longitudinal and thus, no causal conclusions may be drawn. Second, the study findings may not be generalized to hospices outside California. However, the dataset was generated through annual, mandatory reporting required by the State of California. Few states require such mandatory reporting. The results of this study provide a compelling argument for other states to require mandatory reporting of hospice data. Without such data, it is difficult for states to examine services delivered and the services needed to meet the care needs of terminally ill infants and their families. Finally, this analysis was limited to 2008, the year prior to pediatric end-of-life legislative changes in California (Nick Snow Law). Future analyses will need to determine the impact of recent policy changes and the extent to which findings hold over years to come.

Despite these limitations, the findings of this study have implications for hospice clinicians and administrators. The analysis showed that pediatric experience facilitated the delivery of care for infants in nonprofit hospices. Yet, gaining pediatric experience may be a challenge for hospices that have limited exposure to pediatric patients. One possible recommendation for those hospices is to establish a partnership with another health care institution, such as a local children's hospital. Hospices unfamiliar with pediatric hospice care may gain experience when they have a partner on which they can model care practices. Neonatal or pediatric intensive care units, for example, have specialty knowledge of caring for seriously ill children and infants. Additionally, children's hospitals often have experience with a wide variety of end-of-life care delivery models such as inpatient palliative care consultations and pain clinics. ²⁴ These partnerships could take many forms including job shadowing or in-service presentations. In addition, hospices could develop a formal or informal mentoring program with an experienced hospice that delivers pediatric care. Mentoring may provide inexperienced hospices with the opportunity to learn the roles, responsibilities, and routines of caring for children from an organization that has a proven track record in that area. Connecting through visits, e-mails, phone calls, or HIPPA-secure Skype sessions may provide the staff at an inexperienced hospice the opportunity to engage in rounds, case reviews, and clinical documentation. Thus, the pediatric care practices of children's hospitals may be a prototype that hospices with limited pediatric experience can copy.

In summary, understanding the characteristics of organizations is important in the delivery of infant hospice care. Hospices have acknowledged that pediatric care requires specialized knowledge, and many have responded to the need by developing pediatric programs and gaining pediatric experience. Although having a pediatric program in place may demonstrate that a hospice has good intentions to meet the needs of an infant population, this study found that without pediatric experience, hospices were no more likely to provide end-of-life care for infants and their families. The findings of this study support the notion that nonprofit hospices need to explore ways to direct resources toward assisting hospice staff members in gaining pediatric experience.