Reorganizing pediatric rehabilitation services to improve accessibility: do we sacrifice quality?

Children with physical disabilities referred to rehabilitation services can face long delays before receiving these services [1–3]. This situation can affect families' stress, children's psychosocial well being and families' perception of service quality [2, 4, 5]. Despite these consequences, few concrete strategies have been presented in the literature to reduce waiting times for rehabilitation services. Examples include Clow & al. proposing 30-minute rather than 60-minute sessions to children with disabilities [6] and Sorel, Bouchard, Kalubi & Maltais who proposed evaluating children on the waiting lists, counseling their families and offering them information group sessions to foster family well-being while awaiting treatments [5]. Although interesting, these authors did not evaluate the impact of their temporary solutions on the overall quality of service provision (nor on patient outcomes).

Over the years, budget restraints and a growing need for services have led to an increase in delays for rehabilitation in the province of Québec, Canada [3]. Consequently, in September 2006, the pediatric rehabilitation program of the Eastern Townships Rehabilitation Center chose to develop and implement a new model of service delivery to address this issue. At the beginning of the reorganization process, many service providers feared that the strategies proposed to increase accessibility would negatively impact overall service quality [7]. Moreover, reorganization efforts may be destabilizing for service providers and may have a temporary negative impact on service quality [8–10]. Therefore, we conducted a study to determine whether the overall quality of services (as perceived by parents and service providers) was maintained during the reorganization process. We also explored whether child, family, service provider and service-related characteristics influenced these perceptions.

Over the three years, 222 of a possible 904 families participated in the study (2007, n = 69/223, 2008, n = 80/306, 2009 n = 73/375) for an overall participation rate of 24.6%. Analyses performed each year found no differences between responding and non responding families, but respondents tended to have received more hours of services than non respondents in 2008 and 2009. Generally, the characteristics of families were comparable across the years except for their annual revenue (χ² = 6.12, ddl = 2, p = 0.05); a greater percentage of them had a revenue < 35 000 CAN$ in 2008 (Table 1). Children received more total hours of services in 2008 (p = 0.02) and also tended to receive more in 2009 (p = 0.07) compared to 2007, but there were no significant differences in the number of direct hours of services received across the years. In 2008 and 2009, children received more group hours than in 2007 (p = 0.00), but the number of individual hours received did not vary significantly.

Mean MPOC scores ranged from 4.68 to 6.31 and remained stable over the 3 years (p ≥ 0.37 for all domains). The domain Respectful and supportive care received the highest scores and the lowest scores were consistently attributed to Providing general information.

Service providers' consistently rated the overall quality of services (F = 0.52, p = 0.60): 7.59, 7.38 and 7.47 out of 10 in 2007, 2008 and 2009, respectively. Their perception of the overall quality of resources was also stable over time (F = 1.93 and p = 0.15): 6.19, 5.74 and 6.41 out of 10 for the same years.

Bivariate analysis found that overall, only parents' education and child's age influenced the MPOC scores. Families with lower education scored higher on: Providing general information (p = 0.00), Providing specific information (p = 0.03) and Coordinated and Comprehensive Care (p = 0.04). Families with children 12-21 years old scored lower than those with children aged 5 to 11 years for Enabling and partnership (p = 0.03), Providing specific information (p = 0.05), Coordinated and comprehensive care (p = 0.05) and Respect and supportive care (p = 0.02). Hours of services received did not influence the MPOC domain scores (p ≥ 0.22) nor did the type of services (group or not) (p ≥ 0.15).

The service providers' characteristics listed in Table 2 did not generally influence MPOC-SP scores. Younger therapists (p = 0.00) and those with less rehabilitation (p = 0.01) or less pediatric experience (p = 0.03) scored higher on Communicating specific information about the child than those older and with more experience. Full time therapists attributed higher scores than those working part time for Treating people respectfully (p = 0.04) and those who were members of the planning committee scored higher than the others for Providing general information (p = 0.01). Regarding disciplines, special educators attributed higher scores than those of speech and language therapists for Showing interpersonal sensitivity (p = 0.04) and social workers and psychologists attributed higher scores than did occupational therapists (p = 0.02).

With regards to the responses to the open-ended questions, generally the same type of comments came back year after year. Two cross-cutting themes emerged from the codes: collaboration/information exchange and program characteristics. Data regarding these themes provided complementary information to that covered by the MPOC and MPOC-SP items.

Regarding parents' comments on collaboration and information, families appreciated the opportunity to discuss with service providers and to work toward the same goal. Families felt respected and considered the services providers as being competent: 'Service providers are doing what they can with the resources they have and they want our children to progress, they look for the best solutions'. They also appreciated when rehabilitation teams partner with community members and follow up in the child's real life environments. Information exchange appears to be a key point in these partnerships. For example, families were happy not to have to repeat family information over and over again to different service providers. Families also appeared to want more information regarding the program's services (e.g. goals of the interventions and type of existing services) and the availability of community resources (e.g. respite care). Over the years, parents reported appreciating new features to improve information sharing that were introduced by the program, such as encounters for parents while children attend group treatments. Nonetheless, families would appreciate service providers being more proactive in offering information: 'We are often asked about our needs, how are we to know when you don't work in the field? They should make suggestions to us to get us to start thinking about what we might need'.

Services providers also commented on collaboration and information, and discussed some organizational issues. Collaboration with families and community partners was perceived to be important: 'We should involve the parents even more, whenever possible, to counsel more and provide access to good information'. Collaboration was nonetheless sometimes difficult to obtain for different reasons (e.g. families' characteristics, lack of resources or common goals with community partners, school schedule constraints, lack of time, heavy administrative procedures, etc). Community issues were highlighted more often in 2008-2009, probably reflecting an increase in community interventions. During the same timeframe, service providers reported that information sharing with families had improved over the years with the new admission procedures: '[It enables us] to make rapid contact with the client and provide information'. Still, service providers felt that families should be given more time to express their priorities, values and goals. Generally, service providers valued the provision of written information to families and wished to pursue the development of written information tools on services, diagnosis, general counseling, etc. They made suggestions about how to increase the program's efficiency: '[We could] use the center's communication staff to develop information tools for families'. Service providers also valued information exchange among themselves as it prevents duplication of work and fosters case discussion, sharing of intervention techniques and peer support. Moreover, they commented on their own need for information during the reorganization process and wanted quicker responses from their superiors on clinical or administrative issues. Staff turnover in leadership positions created insecurity among service providers and left things unanswered.

With regards to specific characteristics of the program services, many parents commented on the accessibility of services. Indeed, dissatisfaction with waiting times was mentioned year after year. One parent commented the following: 'Services given by intermittence [blocks of therapies interspersed with periods of no treatments] are a way to deal with waiting lists but are detrimental to our children'. Nevertheless, parents appeared to be very satisfied with the quality of services once rehabilitation begins, but many would like more hours of services. A parent reported: 'Compared with the previous year, services are better. However, the duration of services is too short'. Other parents commented on their desire to receive individual therapies and not only groups. A parent said: 'Services should respond to the real needs of the individual child - groups should not be used simply to reduce the waiting lists'. Nonetheless, groups are much appreciated as they break isolation, foster expression and allow parents to observe their children: 'They provide my daughter the opportunity to be with children with similar problems. It is difficult to register her into groups with 'normal' children; teachers are not always patient... Your projects allow us to spend nice moments together!'. In general, over the years, regardless of the type of services received, families reported that services were beneficial to the children as they helped them to become more autonomous, improve their self esteem and quality of life. Many stated that they particularly appreciated when services providers show them ways to work with their children and this helps them deal with daily issues.

Service providers also commented on the specific characteristics of the program services, and more specifically on the different type of services within the new service delivery model, its organizational aspects and the need to carefully plan services based on children's needs. The new admission procedures were seen as allowing them to see children sooner and younger, even if it's not always possible to respond to all of their needs: '[We] get to know everyone on the waiting list and provide different interventions based on needs'. Moreover, the increase in group and community interventions seemed to highlight more the organizational issues around service provision, such as the lack of time and the need for technical support (e.g. having someone to make appointments with families and prepare the materials). Service providers gave different suggestions to be more efficient: 'Simplifying forms, sharing assessment forms (...) to avoid repeating what was already done to be able to spend more time with clients'. Services providers also reported feeling that a lack of resources contributed to limiting their ability to provide many hours of services of different types, often resulting in offering only group interventions to children. A service provider said: '[When only groups are offered], it diminishes the quantity and quality of communication with the parents and thus limits our knowledge about the families and the children'. Logistical issues with groups (e.g. schedule, materials) were also raised by service providers, but one person was encouraged that things were slowly but surely being put into place: 'Things should keep improving over time'.

To our knowledge, this is the first study ever to examine the effect of broad organizational changes on pediatric rehabilitation service quality. MPOC and MPOC-SP scores were not significantly different over the three years, suggesting that the quality of services was likely preserved during the reorganization process. This is encouraging since organizational restructuring has been shown to temporally diminish service quality [8–10]. These results are even more interesting given that the pediatric rehabilitation program internal documents reported an increase of 32% in the number of children receiving rehabilitation services (from 972 in 2007 to 1279 in 2009) and a reduction of 41% in the number of children waiting for a first service (from 95 to 39 for the same period) [24]. Moreover, the number of hours of services received per child increased (total hours of services and group hours), indicating that the reorganization did not reduce the quantity of services provided. Accessibility issues seem to have been successfully addressed without sacrificing service quality.

The observed MPOC scores are similar or slightly superior to those reported in a 2006 review of studies using this tool to measure families' perception of service quality [25]. The lowest MPOC score reported was 3.61 for the Providing general information domain in the context of pediatric services in Britain [26] while the highest reported score was 6.66 for Respectful and supportive care in the context of an early intervention program in the United States [16]. Our MPOC scores are also similar to the only other study involving rehabilitation services in Québec in 1999 [18]. With regards to the MPOC-SP, the observed scores in the present study are generally similar or lower than those reported in the only three other studies using the MPOC-SP [12, 16, 18] presented in the review [25]. A more recent study in Québec used the MPOC-SP and reported similar scores to ours in year 1, although our scores by year 3 were higher than those reported in this other research [2].

Few studies have used the MPOC and MPOC-SP simultaneously. In this research, the MPOC-SP scores were consistently lower than the MPOC scores. This suggests that service providers are perhaps more critical than parents about the quality of care they provide to children and their families (e.g. they have an idea of optimal care whereas parents may not have a comparison base). Dyke et al found similar results [25], while others reported that service providers' scores compared to parents' scores were higher for some domains and lower for others [27, 28]. Only one study using an adapted version of the MPOC with service providers reported service provider scores higher than those of families [18]. The variability in the MPOC and MPOC-SP scores reported is likely due to the broad diversity of settings in which the MPOC has been used.

In general, very few characteristics of the children, families, services and service providers measured in our study influenced the reported scores. These results are corroborated by some findings and conflicting with others in the literature. For instance, Raghavendra et al. (2007) reported that families' level of education influenced MPOC scores and that urban families tended to attribute lower scores than rural families for Providing general information. Moreover, our results showing that families with children 12-21 years old attributed similar scores to those with children 0 to 4 years old differ from those of several studies reporting that families with younger children (often preschoolers) tend to attribute higher scores [21, 26, 28–30]. We believe that our findings, showing only differences for families with children 5-11 years old, might be due to differences in the expectations regarding services as the child grows older [31] or a higher intensity of services provided to younger children [21]. However, we think it could also be due to differences in the services provided by the program before and after entering primary school, and after entering high school.

Service providers' age and experience as well as their discipline, their work status and participation in the planning committee for the service reorganization were found to influence some of the MPOC-SP scores. Interestingly, younger therapists tended to score higher on providing specific information. It is surprising considering that more experience is often associated with higher scores [2, 25]. A possible explanation is the emphasis that has been put on information sharing in academic training in recent years; more recent graduates may have been more sensitized to these issues. The differences shown across disciplines could relate to differences in professional training, while being part of the planning committee exposed service providers to the emphasis that was put on information sharing during the meetings as part of the process of quality improvement.

Regarding service-related characteristics, some have previously argued that the MPOC can differentiate between different forms of service organization (e.g. among families having a care coordinator or not - [32], among children receiving a transition program or not - [26], or among families receiving services from different rehabilitation programs [33]). Until now, the influence of the intensity and the type of service (individual or group) has never been investigated. In our study, these variables were not found to be associated with MPOC scores, despite that many families reported wanting more hours of services and more individual services as well as those provided in a group. These are interesting results as intensity of services is seldom associated with better patients' outcomes [34].

The majority of items that qualified as areas for improvement in the present study are similar to those found by others [25]. Our study is however the first to present longitudinal follow up data for areas for improvement. Interestingly, some areas for improvement disappeared in years two and three, perhaps reflecting the increased awareness by the program of the importance of providing general information to families. Indeed, written materials and new services were developed during the reorganization to increase information sharing. Information kits [35] and having a key worker [36] or care coordinator [26] have been reported as strategies to improve service quality.

This study supports the use of MPOC and MPOC-SP tools to help measure the impact of organizational changes in rehabilitation service delivery. They have been used to evaluate specific modifications in a program or new resources and services [35, 36], but this is the first time that both tools were used to follow up and document perceptions of quality of care during the implementation of a new pediatric rehabilitation service delivery model. Moreover, the additional open opened questions provide useful complementary information regarding themes not included in these tools. These questions raised issues that the program could target to improve service quality (e.g. recommendations regarding organizational issues to increase program's efficiency). One must remember that quality is a multidimensional concept that might be better targeted by the simultaneous use of different tools using mixed methods. The MPOC and MPOC-SP tackle specific aspects of service quality including how services are provided. New types of services such as groups and community interventions might change the dynamic between families and service providers, thus modifying the relationship components upon which the MPOC and the MPOC-SP are built. Future directions to consider in the context of evaluating the perception of the quality of pediatric rehabilitation services should incorporate more service-related variables, such as detailed intensity and type of services measures. Future studies should thus take into account characteristics of service organization and organizational culture. Perception of quality should also be evaluated alongside with a measure of progress of the child to better document the relation between perception of quality and children's outcomes'.

Study limitations include potential recall bias or social desirability, possibly explaining why families tended to score higher than service providers. Moreover, despite various reminders, our response rates for families are among the lowest reported in the literature; two studies report rates below 30% [35, 37], while others have reported rates over 65% [26, 28]. The type and the quantity of rehabilitation services received by the family in the previous year could influence families' willingness to respond. For instance, in our study, non-responding families tended to have received fewer services than respondents thus reducing their desire to participate in the study. Finally, multiple testing may have led to the numerous significant findings.

The results suggest that it is possible to reorganize services to improve accessibility to pediatric rehabilitation services while maintaining service quality. Results also suggest that new service delivery models of pediatric rehabilitation should include features to increase information sharing with families and among interdisciplinary rehabilitation team members. This is especially true during service reorganization when the process can be destabilizing for families and service providers. Understanding and evaluating the multiple dimensions of quality in pediatric rehabilitation services is important to generate useful knowledge to help clinical settings in their continuous quality improvement efforts.

This manuscript is a part of the doctoral thesis of the first author. The thesis is to be published under electronic form in the forthcoming year.