OBJECTIVE: To test the effect of 1 high-intensity, and 1 reduced-intensity, educational intervention designed to improve health care provider attitudes toward youth with sickle cell disease (SCD).
METHODS: We exposed a regional sample of pediatric health care providers to a 2.5-day high-intensity educational and experiential intervention using videos about the SCD patient experience. Additionally, we traveled to a different set of regional health care institutions and offered pediatric providers a reduced-intensity intervention, consisting of a 90-minute lunchtime in-service centered on our same set of videos about the patient’s experience. We assessed the impact of both interventions by taking pre/post measurements of the negative and positive attitudes expressed by participating providers toward patients with SCD.
RESULTS: Both interventions tested elicited improvements in the SCD attitudes expressed by the pediatric providers as suggested through a reduction in measured negative attitude scores (20.0 vs 12.1, P < .001), and an improvement in positive attitude scores (67.1 vs 72.2, P < .001). Further testing suggested that the high-intensity intervention elicited a stronger effect than the reduced-intensity intervention across multiple attitudinal domains.
CONCLUSIONS: Video-based interventions can be used to improve the attitudes of pediatric providers toward patients with SCD. The availability of interventions of varying intensities provides greater flexibility in designing efforts to advance the quality of SCD care through the improvement of provider attitudes.
Health care provider attitudes toward certain patient populations are known to affect health care quality. Negative health care provider attitudes can lead to a number of poor clinical and patient-reported outcomes, and can contribute to the overall problem of health care disparities.1–3 Physicians with more negative attitudes about certain, often stigmatized, patient subgroups have been found to engage in worse perceived communication with patients, less rapport-building with patients, less information-giving to patients, worse estimated ratings of the patient’s level of adherence, worse patient satisfaction with the visit, and worse patient-reported health status.4–7 The identification and elimination of negative health care provider attitudes is of instrumental importance because of their known association with poor patient outcomes. Similarly, the study of these attitudes is of intrinsic importance because they are departures from the ethical mandate health care providers have to show respect for patients as unique persons, which is a fundamental requirement of social justice and the principles of medical ethics.9,10,12
Sickle cell disease (SCD) is one condition in which negative health care provider attitudes are known to be prevalent and to affect the SCD patient-physician relationship.15 In the United States, African Americans experience the highest burden of the disease.16 Persons with SCD experience significant morbidity throughout their lives in the form of severe acute and chronic pain, progressive organ and tissue damage, susceptibility to infections, strokes, and neurocognitive impairment. Persons with SCD have higher levels of acute care utilization, and the median life expectancy of persons with the most severe form of the disease is reduced by 30 years compared with the general population.17–19 Although the severity of the condition and the intense recurring pain it causes are hallmark symptoms of SCD, pain management for the disease is thought by both patients and health care providers to be inadequate.20–22 A systematic review of potential barriers to SCD care found substantial evidence to suggest that negative health care provider attitudes toward patients with SCD serve as a significant barrier to the delivery of appropriate pain management.15 Examples of these negative attitudes are found in reports of adult patients with SCD that health care providers fail to show them respect by stigmatizing them as being “drug-seeking,” that providers often lack trust in them by expressing doubts about the veracity of the patient’s reports of their pain experience, and that providers do not treat the patient with SCD with a caring approach.15,23 Although formal studies of these attitudes and experiences among youth with SCD are much more limited in number, the studies that can be identified document many of the same issues,24 and they comport with the experiences of youth with SCD who we talk to and care for in our own professional practices. Furthermore, parents of hospitalized children with SCD report many of the same difficulties in their experiences and relationships with health care providers as reported by adult patients. For example, parents of these young patients have been found to be more likely than parents of children hospitalized with other conditions to report an inadequate level of involvement in decision-making about their child’s care during the hospital stay. This result persisted even after controlling for the race of the parents.25,26
In an effort to address the problem of provider attitudes toward patients with SCD, our group previously reported on a novel approach that uses videos of patients with SCD and experts in SCD care discussing the problems that many of these patients face in seeking care.27 In this study, we found that internal medicine providers exposed to our video intervention exhibited more positive attitudes about patients with SCD than a comparison group of providers not exposed to the video.27 Although our previous work is significant in demonstrating that provider attitudes toward patients with SCD can be improved, our previous findings are limited in that the providers under study cared for only adults with SCD. Little is known about the attitudes and beliefs that pediatric health care providers have toward the SCD population, and the extent to which these attitudes are amenable to change. The current study addresses this important gap. We developed 2 different, variable-intensity, video-based interventions designed to educate health care providers about SCD and to improve the attitudes toward patients with SCD expressed by pediatric providers. The aims of the study reported here were (1) to identify any impact of the interventions (in the aggregate) on pediatric provider attitudes toward patients with SCD, and (2) to compare the respective magnitudes of any impact attributed to the interventions separately. We hypothesized a priori that the interventions viewed in the aggregate would be successful in improving provider attitudes, and that we would observe a larger magnitude of effect from the high-intensity intervention compared with the reduced-intensity intervention. We present these impact data for the benefit of institutions and educators interested in using our materials as part of their own institutional educational efforts regarding SCD. Our materials are accessible at http://www.bioethicsinstitute.org/research/projects-2/hope.
The current study is 1 component of a larger federally funded effort designed to increase the integration of the principles of pediatric palliative care into the ongoing care of adolescents living with life-threatening chronic diseases. The study reported here uses a quasi-experimental multiple-group pretest-posttest design. The research procedures described were reviewed and approved by the funded institution’s institutional review board.
High-Intensity Intervention (the “Retreat”)
As we tested 2 different interventions (ie, 1 high-intensity and 1 reduced-intensity), there were 2 general components to our recruitment approach. For the high-intensity intervention (ie, “the retreat”), our participants were multidisciplinary teams of health care providers from a selected sample of health care institutions located east of the Mississippi and representing north, mid-, or southern regions of the United States. Institutions were potentially eligible if they had an established sickle cell clinic that served youth between 13 and 21 years of age, and if they had a pediatric or adult palliative care program that had not previously been a recipient of, or otherwise exposed to, the educational curriculum developed by the Initiative for Pediatric Palliative Care.28 After identifying a list of potentially eligible institutions, we contacted individuals at the institutions who could serve as a “site advocate” that would champion participation in our study among institutional staff, and who would be willing and able to coordinate the recruitment of individuals from their institution into our study (by using a combination of personal communication, print, and electronic advertisement). We identified these potentially eligible institutions and potential site advocates through our own professional networks and contacts, and with assistance from members of an external advisory board that we convened to aid in study oversight. The multidisciplinary teams of clinicians invited to participate included physicians (pediatric hematology/oncology, emergency medicine, and palliative care fellows), nurses (pediatric inpatient, adolescent inpatient, adolescent medicine, palliative care, pediatric emergency, sickle cell clinic, and palliative care nurses), palliative care specialists, clinical psychologists, social workers, genetic counselors, and other specialists.
Reduced-Intensity Intervention (the “In-Service”)
We used the criteria and procedures described to select eligible institutions, identify site advocates, and to recruit participants into our reduced-intensity intervention. Although the types of clinicians participating in the reduced-intensity intervention were similar to those listed for the high-intensity intervention, physicians comprised a larger proportion of the reduced-intensity intervention participants than were found among the high-intensity intervention (Table 1, 34% vs 14%, P = .035), with the reduced-intensity intervention physicians tending to be at an earlier stage in their careers.
The High-Intensity Intervention
We developed 2 interventions of varying intensities that were designed to educate clinicians about palliative care, SCD care, and approaches to their integration.29 Adapted from the program developed by the Initiative for Pediatric Palliative Care, the high-intensity intervention (ie, the retreat model) consisted of the development of a comprehensive curriculum on palliative care, SCD, and the unmet needs of persons living with SCD, as well as the potential for integrating palliative care principles into SCD care. The material was structured around 6 documentary videos that covered topics such as the physical and psychosocial impacts of SCD on well-being and quality of life, challenges in the provision of appropriate pain management, the building and breaking of trust in the SCD patient-clinician relationship, challenges in transitioning from pediatric to adult SCD care, and the moral distress experienced by clinicians whose careers focus in large part on caring for persons with SCD. The videos were composed of interviews with patients living with SCD, their families, and SCD clinician experts. All curricular materials were presented by using the format of a 2.5-day-long educational retreat held in Annapolis, MD. The retreat was composed of various educational and experiential activities, including didactic presentations by the primary investigators and small-group discussions based on the videos that were facilitated by research team members and trained community experts. We administered a comprehensive pre/post retreat survey that included a measure of health care provider beliefs and attitudes about patients with SCD (described in more detail later in this article).
The Reduced-Intensity Intervention
For our reduced-intensity, in-service intervention, we adapted our 2.5-day-long retreat into a 90-minute version more suitable for use during a typical “grand rounds” or “lunchtime seminar” format. With this model, 2 of our primary investigators traveled to the participating institutions where they showed our educational videos to an assembled audience of pediatric health care providers, and then led members of the audience in a short debriefing discussion about what was learned and what can be incorporated into provider practices going forward. Selected measures from our attitudinal questionnaire were administered to the participants before and after the session. We held our in-service sessions during preexisting or specially scheduled events that the institutional site advocates coordinated.
Health Care Provider Attitudes
Health care provider attitudes toward persons with SCD served as the outcome measures for this study. We assessed these attitudes by using our previously validated instrument, the “General Perceptions About Sickle Cell Disease Patients Scale.” Developed during our previous study,27 this measure assesses 4 dimensions of provider attitudes. The negative attitude subscale assesses the extent to which providers exhibit negative attitude toward patients with SCD (ie, a measure of disrespect), which includes health care provider perceptions of the extent to which patients with SCD are likely to abuse or be addicted to opioids, alcohol, and illicit substances; the extent to which patients with SCD attempt to manipulate health care providers or otherwise exaggerate the existence or severity of their pain; and the extent to which patients with SCD are likely to comply with medical recommendations. The positive attitude subscale assesses the extent to which providers exhibit more positive feelings of affiliation for patients with SCD. The concern-raising and red-flag behaviors subscales both measure the extent to which certain behaviors exhibited by patients with SCD are signs that the patient is inappropriately drug-seeking (ie, measures of trust). Scores on these subscales range from 0 to 100, with higher scores indicating a greater amount of the dimension measured by the subscale (eg, higher scores on the negative attitudes subscale signify more negative attitudes/disrespect toward patients with SCD, and so forth). These measures have demonstrated good psychometric properties among health care providers in previous studies, and the instrument has subsequently been used or adapted for use in a number of studies examining provider attitudes toward patients with SCD.30,31
Comparing Pre/Post Intervention Attitudes
We administered our survey instrument to participants in both of our interventions immediately before commencement of the intervention (ie, at the sites of the retreat and in-services, but just before commencement of the respective intervention), and again immediately after completion of both interventions.
To meet the first aim of this study, we calculated and report pre/post intervention scores on each of the attitudinal subscales comprising our outcome measures. We report unstandardized measures of the effect size (ie, the difference in postintervention and preintervention mean attitudes scores) for our interventions in the aggregate. Bivariate (unadjusted) analyses for this aim were conducted using the paired t test, and multivariable (adjusted) analyses for this aim were conducted by using linear regression within a generalized estimating equations framework. For the second aim, we report the unstandardized effect sizes (ESs) observed for each intervention separately, and we compare the ESs observed for the high-intensity intervention with that of the reduced-intensity intervention. Paired t tests were used to conduct unadjusted tests of the significance of any effect observed for each intervention separately, and linear regression models testing for effect-modification within a generalized estimating equations framework were used to test the relative magnitudes of each intervention’s impact. We included health care provider gender and race as covariates in all regression models because of the association of these characteristics with provider attitudes toward SCD found in our previous work,30 and among our preliminary, unpublished data. We compared the respective magnitudes of ESs from our different interventions by constructing regression models with interaction terms to test for the presence of effect-modification by intensity of the intervention (retreat versus in-service). A statistically significant interaction term suggests that the magnitude of the intervention effect differs comparing the retreat with the in-service. An α of 0.05 was used as the threshold for statistical significance. The Stata 13.1 statistical software package was used for all statistical analyses (Stata Corp, College Station, TX).32
We present data on 103 health care providers who participated in our in-service intervention, and 43 who participated in our retreat. Table 1 provides the characteristics of our participants by intervention type. Women and whites comprised most participants for both interventions. Health care teams from 3 states (Ohio, New York, and North Carolina) participated in our in-service, whereas teams from 4 states participated in our retreat (Michigan, Ohio, Pennsylvania, and South Carolina).
Testing the Intervention Effect and the Relative Impact of the Retreat Versus In-Service Models
Tables 2 through 4 summarize the observed scores on each attitudinal measure overall, and by intervention type. Table 5 shows the results of unadjusted statistical analyses that assess whether the combined effect of our 2 interventions exhibited a significant impact on pediatric health care provider attitudes, as well as showing the separate effects of each intervention. As shown in the table, the pooled impact of the interventions significantly improved pediatric provider attitudes in each of the 4 attitudinal dimensions. As illustrated by the reported ESs, postintervention scores for the negative (ES –11.9), concern-raising behavior (ES –18.7), and red-flag behavior attitude scores (ES –19.1) were each decreased compared with preintervention scores for these measures (all P values < .01), whereas postintervention scores for the positive attitudes increased compared with the preintervention measure (ES 12.7, P < .01). In analyses by intervention, we found that each of the 4 attitudinal scores showed improvement after the retreat intervention. For the in-service intervention, however, postintervention scores improved for the negative, concern-raising behavior, and red-flag attitudes only. The difference in post/preintervention positive attitude scores for the in-service did not reach the threshold for statistical significance used in this study (ES 2.5, P = .08).
The pooled intervention results for analyses adjusted for the gender and race of the provider were similar in direction to the pooled results reported previously for the unadjusted analyses (Table 6). Specifically, negative attitudes (β = –8.16, 95% CI –10.11 to –6.21), concern-raising behavior beliefs (β = –14.76, 95% CI –17.76 to –11.76), and red-flag behavior beliefs (β = –13.54, 95% CI –16.53 to –10.56) were each decreased in the pooled postintervention assessment compared with the preintervention assessment, whereas positive attitudes were increased in the postintervention assessment compared with the preintervention assessment (β = 5.27, 95% CI 2.35 to 8.19).
Table 7 shows the effect of our intervention stratified by retreat/in-service status. The P values listed in Table 7 show the statistical significance of our interaction term that tested whether there were differences in the magnitude of any observed intervention effects based on the type of intervention used, while still adjusting for the gender and race of the provider. For 3 of the 4 attitudinal measures, the retreat version of the intervention exhibited a statistically significantly stronger impact on pediatric provider attitudes than the in-service intervention. For the positive attitudes subscale, the in-service intervention appeared to have no statistically significant effect on pediatric provider attitudes (βIn-service = 2.60, 95% CI –0.31 to 5.51), whereas positive attitudes did significantly increase during the retreat intervention (βRetreat = 12.7, 95% CI 5.72 to 19.66). Although the in-service intervention decreased negative attitudes and red-flag behavior beliefs, the retreat intervention exhibited a higher-magnitude decrease in these dimensions. No statistical difference in the magnitude of the intervention effects was observed for the concern-raising behavior beliefs, as both intervention models decreased these beliefs among the pediatric providers with a similar magnitude (βIn-service = –13.04, 95% CI –16.26 to –9.83; βRetreat = –19.4, 95% CI –26.17 to –12.65).
We conducted a study among pediatric health care providers that tested the impact of 2 variable-intensity video-based educational interventions designed to improve attitudes toward patients with SCD. As in previous work, we found that provider attitudes can be improved by way of using novel video-based interventions that showcase persons with SCD and expert SCD providers discussing challenges and strategies in caring for persons with SCD. We also found that the intensity of the interventional approach matters, and that high-intensity interventions may be used to observe stronger improvements in provider attitudes.
Overall, we found that our interventions, when considered together, were shown to improve pediatric provider attitudes across all 4 of the attitudinal dimensions that we studied. A similar finding was observed in our previous study that was shown to improve attitudes in these same 4 dimensions among adult health care providers.27
We further found that the intensity of our interventional approach made a difference in the magnitude of the observed effects. Our high-intensity “retreat model,” which consisted of a 2.5-day-long workshoplike session in which health care teams were secluded together for a series of small-group and plenary discussion sections, was observed to elicit a greater improvement in the negative attitudes and the red-flag behavior beliefs exhibited by participating providers than was observed among those participating in our reduced-intensity “in-service model.” Further, the retreat model was found to increase the positive feelings of affiliation exhibited by the health care providers toward patients with SCD, whereas the in-service model was not observed to have any effect on positive attitudes. This suggests that the high-intensity approach may be the preferred approach to improving attitudes in those situations in which baseline provider attitudes toward patients with SCD are thought to be particularly bad, such as in situations in which historically there has been a great deal of conflict between the local patient population with SCD and health care providers at a particular institution.
Nevertheless, our reduced-intensity in-service model intervention was shown to elicit improvement in negative attitudes and red-flag behavior beliefs, and there was no difference in the observed improvement in concern-raising behavior beliefs comparing the retreat to the in-service models. This suggests that our reduced-intensity intervention can be used on its own or in conjunction with other approaches to improve provider attitudes toward patients with SCD. Compared with the retreat model, the reduced-intensity in-service model has the benefit of being less costly to implement, it can be used to serve a greater number of health care providers in 1 session than the retreat model, and the in-service model is a more natural fit with activities that typically are used in health care institutions to provide continuing medical education. Furthermore, the in-service model can more readily be adopted for use in Web-based interventions, which could greatly expand the reach of these interventions on a regional, national, or even international basis.
Some important limitations of our study must be considered. Because some of the participants in our retreat intervention were invited because of their employment at the specialty SCD centers at their institutions, these retreat participants might be considered by some as a self-selected sample of health care providers who were previously invested in opportunities to improve the quality of care delivered to patients with SCD. If true, this could lead to an overestimate of the true effect of our intervention, and it could raise questions about the generalizability of our findings. Nevertheless, our in-service participant sample was larger than the retreat sample, and we estimate that these participants were less likely to be members of their institution’s SCD clinic. Despite these differences, we still observed significant improvements in attitudes from this interventional model. A second limitation of our study is the still poorly defined link between attitudes toward patients with SCD as assessed by this measure and actual provider behaviors toward these patients. The extent to which changes in attitude scores observed using this measure are associated with measurable changes in the quality of care provided to pediatric patients with SCD is unknown. That is, the relationships between observed statistical ESs and the clinical significance of varying ESs has yet to be established. A previous study has established that emergency providers with the highest scores on the negative attitudes subscale were less likely to follow recommendations for the readministration of pain medicines after an initially ineffective dose, thus establishing an initial link between attitudes and actual clinical behaviors. Similar work is needed for other adult health care providers when using this measure, as well as for pediatric providers. Finally, the collection of eligibility requirements developed for this study might provide some with concern regarding the generalizability of our findings. It is important to note, though, that the findings of this study are consistent with our previous attempts at using similar video-based interventions to improve provider attitudes among an increasingly wider selection of provider types (eg, adult emergency providers, adult internists), which suggests that the core components of our approach may provide some beneficial impact across a wide range of settings.
Health care provider attitudes about certain groups of people are known to affect the quality of care delivered to patients. It is well-documented that many health care providers exhibit negative attitudes toward patients with SCD. Nevertheless, we found that our video-based interventions were able to elicit improvements in the attitudes expressed toward this patient population among a group of pediatric providers. Even though our intervention models were of varying intensities, they both were shown to have beneficial effects on attitudes. This makes our interventions suitable for use in a variety of different strategies designed to improve the quality of SCD care by affecting the attitudes of providers toward persons with the disease. Improving the quality of SCD care has been identified as a high priority goal of US health policy efforts.32 The results of this study suggest 2 more strategies that are worthy of consideration in pursuit of this greater goal, so that ultimately persons and families affected by SCD will be more likely to receive the high quality of care, and are treated with the appropriate levels of respect and trust, that all patients are due.
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
FUNDING: All phases of this study were supported by a National Institutes of Nursing Research grant 1RC1NR011710-01. Supplemental support for the study was provided by the Stavros Niarchos Foundation. Dr Haywood’s effort was funded by a National Heart, Lung, and Blood Institute Career Development Award 1K01HL108832-01. Funded by the National Institutes of Health (NIH).
POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.
- Smedley BD,
- Stith AY,
- Nelson AR
- Powers M,
- Faden RR
- Beauchamp TL,
- Childress JF
- Kaye CI,
- Accurso F,
- La Franchi S,
- et al
- ↵National Heart, Lung, and Blood Institute. Workshop on adults with sickle cell diseases: Meeting unmet needs. June 13-14, 2002. Bethesda, MD.
- Maxwell K,
- Streetly A,
- Bevan D
- ↵Initiative for Pediatric Palliative Care. Home page. Available at: www.ippcweb.org. 2014. Accessed March 19, 2014.
- Rushton CH,
- Geller G,
- Caldwell M
- ↵Office of Minority Health. Sickle cell disease: increasing access and improving care. Available at: http://minorityhealth.hhs.gov/assets/pdf/Checked/1/sickel_cell_anemia_factsheet.pdf. Updated 2011. Accessed April 12, 2015
- Copyright © 2015 by the American Academy of Pediatrics