This study identified factors affecting health care expenditures and magnitudes of their associations for Local Government Medical Benefit Scheme (LGMBS). Specifically, the study compared population structure and health utilization pattern between LGMBS and Civil Servant Medical Benefit Scheme (CSMBS) that provides a comparable benefit package with a relatively higher expenditure. Based on data in fiscal years 2014-2015, major findings are as follows:
In 2015, persons eligible and registered to LGMBS amounted to approximately 0.6 million, a 6% increase from that in 2014. Comparing to CSMBS with 4.8 million members, LGMBS had a lower proportion in male (44 vs. 47%) and elderly people (26 vs. 33%). The LGMBS members in each age-group were less likely to be hospitalized in a year, as compared with the same-age CSMBS members (For female-male: children 6-8 vs. 10-11%; adults 7-5 vs. 7-6%; elderly 14-15 vs. 18-15%). Such a difference in the demographic profiles resulted in a lower patient proportion of the LGMBS than the CSMBS (all hospital types: 8.8 vs. 9.4%).
The second factor is socio-economic status. Members of LGMBS tend to have a lower education and living standards than CSMBS (Higher education 26 vs. 42%; Living in the top 20% richest households –quintile 5 26 vs. 50%). The key factor leading to choices of service use is resident areas, whereby 30% of the LGMBS members resided inside the municipality area and 58% of the CSMBS did so. Patients visiting health facilities with a higher and expensive care level such as university hospitals were lower in LGMBS than CSMBS (Outpatient, OP: 15-16 vs. 25-26%; Inpatient, IP: 9 vs. 15%).
After controlling for differences between the two Schemes in various characteristics, including hospitalization years, patient sex, age and disease severity, and health zone of hospitals, the LGMBS patients were less likely to be admitted to a provincial or university hospital (PH or UH), which is larger than a district hospital (DH) than the CSMBS (P<0.05) for 8 in 10 most common hospitalized illnesses by a magnitude ranging from 14% for pneumonia to 24% for diabetes (odds ratio, OR: 0.76-0.86).
Difference in the ages also led to a lower frequency of health utilization in the LGMBS patients (OP visits via direct disbursement: PH&UH 4.9-5.3 times and DH 4.6-4.9 times vs. 6.1-6.2 times; IP admissions: 1.5 vs. 1.6 times). For severity of the diseases leading to hospitalization, the LGMBS patients tend to have a lower proportion of severe-very severe levels than the CSMBS (Top 50-diseases: 19 vs. 24% and the rest: 17 vs. 22%). Such a difference could lead to a relatively shorter stay for LGMBS in all-type hospitals (6.9-7.0 vs. 9.0-9.1 days) and in each hospital sub-type (DH 4.4 vs. 5.1 days; PH 7.0 vs. 9.0 days; UH 8.7-9.1 vs. 10.4-10.7 days). Once hospitalization years, patient sex, age and disease severity, hospital types, health zone, and disease groups have been accounted for, the LGMBS patients stayed in hospitals shorter than the CSMBS by 0.32 days (P<0.01) for all-type hospitals and ranging from by 0.17 days in DH to 0.36 days in PH and 0.39 days in UH.
Demographic, socio-economic and residential characteristics plus clinical conditions stated previously are key factors that made the LGMBS expenditure per patient or per visit in 2014-2015 lower that the CSMBS (OP via direct disbursement: 6,019-7,019 Baht per patient or 1,092-1,170 Baht per visit vs. 1,667-1,623 Baht per visit; IP: 28,495-30,354 vs. 41,299-42,852 Baht per patient or 18,903-19,742 vs. 26,658-27,556 Baht per admission; and for adjusted RW: 2.1-2.2 vs. 2.7-2.7 units per patient or 1.4-1.4 vs. 1.7-1.7 units per admission). Furthermore, a significant number of LGMBS OP patients (30-34%) were not found to have a payment claim.After controlling for patient characteristics, hospitalization years, hospital types and health zones; the LGMBS patients had a slightly higher adjusted RW than the CSMBS for all-type diseases (by 0.018 units in all-type hospitals and by 0.067 units in UH). However, the LGMBS’ IP expenditure was relatively lower (by 529 Baht per admission in all-type hospitals, by 375 Baht per admission in DH and by 752 Baht per admission in PH). Specific to diseases, the LGMBS patients had a relatively lower adjusted RW (by 0.025 units in DH for bronchitis; by 1.19 units in UH for pneumonia; by 0.006 units for breast cancer with chemotherapy; and by 0.005 units for lung cancer without chemotherapy). The IP expenditure for LGMBS was also lower than CSMBS for all top-10 common diseases (diarrhea, pneumonia, diabetes, bronchitis, COPD, CKD, CHF, CVA, UTI and appendicitis) plus cervical cancer.
The facts that health expenditures for LGMBS were lower than CSMBS even after controlling for demographic and clinical characteristics of patients as well as hospital characteristics hint to additional factors remaining unaccounted for due to unavailability in analytic datasets. These could include socio-economic status, residential areas and choices of health technology such as use of expensive, non-essential medicines, which were found a higher use by the CSMBS patients in all hospital types. In addition, efficiency in the management system by National Health Security Office (NHSO) through the diagnostic related group (DRG) for IP and fee-for-service payment for OP services through direct disbursement and receipt systems.
Results from a model simulation predicting the LGMBS budget increase due to selective scenarios assuming the LGMBS patient characteristics comparable to the CSMBS as follows: (1) same age-sex structure: the LGMBS expenditure would increase by 14% of the current 2015-OP expenditure (2,566 million Baht) and by 29% of IP expenditure (1,656 million Baht) over the same year or in total by 830 million Baht (combined OP-IP); (2) same age-sex specific expenditure per patient: the LGMBS expenditure would increase by 29% and 10% of the current OP and IP expenditures, respectively or by 920 million Baht in total; and (3) assuming both (1) and (2): the LGMBS would increase by 45% and 41%, respectively or by 1,850 million Baht in total.
Results from the qualitative study using in-depth interviews with key partners, including patients, the LGO’s staff and health care providers revealed findings consistent with the quantitative analysis that due to a relatively young age of the LGMBS members, they were relatively healthier. Once they got sick, a health facility closed to their homes, whereby in provincial areas is typically a DH or health center is a usual source of services. Management by NHSO including for example reimbursement audit could reduce duplicating enrollment and rational drug use according to treatment guidelines. In addition, existing cost-containment measures on the use of non-essential medicines that have been implemented for CSMBS patients had a spill-over effect on the LGMBS.
Service users who were first-line relatives of the LGMBS beneficiaries may not be aware of eligible benefits or may opt out by paying out of their own pockets (OOP) or using private insurance. Results from an additional survey of the LGMBS patients found that 59% of users of private health facilities have paid OOP with the median utilization of 3 times in a year.
For perception on the services obtained, both LGMBS patients and providers did not distinguish the quality of care to be poorer than the CSMBS. In addition, they agreed certain patients did not understand about the eligible benefits.
Perceived improvement in the NHSO performance on managing the LGMBS included time taken for reimbursement through receipt system, co-ordination made by the NHSO Regional Office, member registration during work transition, direct disbursement to selected health centers, claim processing for selective benefits such as expensive medicines.