Health care disparities in timing of diagnosis and continuity of care among children with cerebral palsy under Taiwan's national health insurance program: A population-based longitudinal study

Background: Taiwan provides universal health insurance coverage for almost 98% of the population since March 1995. One of the objectives of this national health insurance program is to achieve equitable access to health care. Previous studies related to health care disparities mainly focused on the utilization of health care itself, however, little study addressed socioeconomic or geographic disparities in early diagnosis and continuity of care. We used a population-based longitudinal data set to investigate health care disparities among children with cerebral palsy (CP) under Taiwan's national health insurance (NHI) program. There are two objectives of this study: (1) to understand the timing of diagnosis of CP and factors associated with early diagnosis of CP under Taiwan's NHI program; (2) to report the continuity of CP outpatient services use and outpatient therapy treatments for a 3-year period and investigate factors associated with continuity of care for these CP children.

Methods: Those who were born between 1997 and 2005 were retrieved from the NHI clams database. Based on this population data, we identified cerebral palsy (CP) children with three or more CP ICD9 codes (343.X or A224) before year 2005. Data files for these CP children include enrollment data, outpatient records, physician and medical institution files, as well as medical orders prescribed. Variables created from these data set include patients' demographic information (age, gender) and their socioeconomic factors (parents' income and employment level, the geographic and urbanization level of residential area), timing of diagnosis, characteristics of medical setting (level and urbanization of location), physicians' age, specialty and experience with CP patients, utilization of well-baby care, utilization of CP outpatient services, physical therapy, occupational therapy and speech therapy. For the first part of our analysis, we dichotomized the timing of diagnosis into before and after 1.5 years old. Logistic regression model was used to investigate socioeconomic or geographic disparities in early diagnosis with CP. For the second part of our analysis, we used two measures for continuity concept in this study. The first measure was created by categorizing CP children into different types based on their continuity of CP-related outpatient use as well as their outpatient therapy treatments within three years. We used multinomial logit regression model to investigate factors associated with their continuity of use. The second measure was created by calculating the number of CP-related outpatient services and outpatient therapy treatments. We used negative binomial regression model to analyze factors associated with the number of CP-related outpatient visits and outpatient therapy treatments.

Results: Of the 1.68 million children born between 1997 and 2005, 8,034 children were diagnosed with CP. The average age at first diagnosis was 1.57 years old for these birth cohorts. In the oldest birth cohort, the cumulative incidence rate up to age 9 was 4.8 per thousand. The age-specific incidence rates of CP were homogenous across birth cohort, with peak around 1.5 years old. For those can be observed for at least three years, their average number of CP-related outpatient visits for the first three years after diagnosis were around 13 to 14 times and their correspondent average expenditure were around 1,005 to 1,238 US dollars. Our multivariate logistic regression analysis showed that children from the highest level income family were more likely to be diagnosed earlier. No urban-rural disparities were found to be associated with timing of diagnosis. As for the continuity of CP-related outpatient services, for those who can be observed at least for three years, only 65% of CP children continued to use CP-related outpatient services in the second year after being diagnosed and only less than 60% would continue to use CP-related outpatient services in the third year. Around 89% of CP children would use at least one type of outpatient therapies in the first year after being diagnosed, however, the utilization rate dropped to 60% in the second year and declined to 55% in the third year. For three types of outpatient therapies (physical therapy, occupational therapy and speech therapy), the utilization rate in the first year after being diagnosed were 85% ,67%, and 23% respectively. In overall, the utilization rate declined for physical and occupational therapy but increased for speech therapy. In the second year after diagnosis, the utilization rate dropped by 32% and 24% for physical therapy and occupational therapy respectively, however, the utilization rate increased by 17% for speech therapy. In the third year, the utilization rates were 53%, 48% and 29% for three types of therapies. Around 4 to 10% of CP children would continue to use any type of therapies in the second year and 10 to 23% would use any type of therapies 1 or 2 years after being diagnosed. Around 50% of CP patient have never used speech therapy in the first three years with first-diagnosed age at 3 or greater than 3. Based on our multinomial logit model, those from highest level of income family were less likely to not use any CP-related outpatient service 1 year after being diagnosed. In addition, our study also showed that those from highest level of income family were also less likely to not continue using any outpatient therapy services in three years after being diagnosed with CP. In terms of the frequency of use, results from negative binomial regression model showed that those from highest level of income family were more likely to visit more to CP-related outpatient services, and those who resided in rural area were less likely to have more outpatient therapy visits than those who lived in urban area.

Conclusion: We investigated health care disparities in terms of the timing of diagnosis with CP, frequency of CP-related outpatient services and physical therapy treatments as well as the continuity of outpatient and therapy treatments. Our findings suggest that even though Taiwan has provided universal insurance coverage, health care disparities still existed among socioeconomic groups. Moreover, the frequency of visit to three outpatient therapies varied between urban and rural areas, suggesting that the accessibility to these services in rural areas deserve more attention. National Health Insurance coverage itself cannot make equal access to health care possible alone, it needs an integrated healthcare and social welfare delivery system in order to provide equitable accessibility to health care for children with special health care needs. In addition to the supply side of health care providers, intervention strategies from the demand side should be targeted on children in low income families including the promotion of the concept on early diagnosis of diseases, and the importance of the continuity of care for children with chronic conditions. Finally, the policymakers should also examine the current free well-baby care in order to make it more cost-effective in playing the role of early diagnosis of disease and providing medical consultation to parents or caregivers with CP children. Gatekeeper system could be established based on this current service in order to transfer those children with special health care needs to appropriate healthcare facilities for treatment and also assure their continuity of care.