Determinants of Clinical Practice Variations and Influence of Provider Payment Methods: A Case Study from Thailand

Abstract:     
Rationale: In 2001, Thailand achieved universal coverage (UC) in access to health care for the whole population by introducing a tax-financed health insurance scheme to 47 million people (~75% of total population). The UC scheme employs a capitation payment for ambulatory care and a global budget with diagnostic-related group (DRG) for hospitalization. A nominal payment of 30 Baht (~USD 0.75) per ambulatory visit or hospital admission was levied. In rural areas, district health system (DHS) is the main contractor whereby the UC members are required to register with the nearby DHS. In urban areas, public and private hospitals are the main UC contractors. A benefit package of the UC scheme covers outpatient care, hospital admissions, health promotion and disease prevention, as well as a wide range of high-cost care with an additional fee schedule outside the capitation and global budget.

The Social Security Scheme (SSS) is a compulsory social health insurance financed by tripartite contributions for 9 million private employees in the formal sector. Payment to health providers rely on a pure capitation basis.

The Civil Servant Medical Benefit Scheme (CSMBS) is a non-contributory, tax-funded scheme with free access to a comprehensive service package, but limited to public hospitals without co-payment. Healthcare providers are paid on a conventional fee for service. This results in rapid cost escalation. In 2006, the expenditure per CSMBS member, 5,300 Baht (USD143) was 3.2 times that of per UC member, 1,654 Baht (USD48). Policy makers need to be informed if cost containment from the capitation payment would result in poor quality and welfare loss, and vice versa, if expensive CSMBS would lead to higher level of quality and survival.

Objectives: In the context of the close-end payment of UC, the capitation payment of SSS, and the open-ended payment (fee-for-service) of CSMBS, this study explores variations in clinical practice, costs, and outcomes of selected medical interventions provided to members of these three health insurance schemes, in order to inform policy decision if there is any need to fine-tuning payment methods to achieve equity and efficiency objectives.

Methods: This study analyzes nationally patient databases of three health interventions provided to UC, SSS, and CSMBS members, and employing literature review to compare costs, intermediate and clinical outcomes among the three different schemes.

Results: Evidence from approximately 1.2 million deliveries in hospitals during 2004-2006 indicates that pattern of the deliveries varied substantially across health insurance schemes. There existed an increasing trend in Caesarian section over time. Women covered by the CSMBS underwent the Caesarian section in a much greater proportion than the SSS and UC members. Provincial hospitals performed this expensive procedure more frequently than district hospitals and the rest. The CSMBS members have been admitted to the provincial hospitals for the delivery purpose more often than in the district hospitals when compared with the SSS and UC members.

The UC scheme lists leukemia, a life-threatening hematological cancer, as a high-cost care with a separate payment schedule. Empirical evidence from 2003-2005 patient registry indicated more effective but expensive chemotherapies were given to the acute non-lymphoid leukemia (ANLL) patients who were covered by SSS and CSMBS in a greater proportion than UC members. This resulted in better survivals for SSS and CSMBS beneficiaries than for UC members when adjusting for age difference.

In 2000-2002, prescribing variation for chronic asthma patients (N=6,176) was observed in 18 provincial hospitals. The UC members who were exempted from the 30-Baht co-pay and SSS members were less likely to get the steroid inhaler treatment which is deemed necessary for asthma control than the CSMBS members even though the underlying difference in asthma severity was taken into account.

Conclusions: Determinants of clinical practice variations are complex, and provider payment is one determinant of such variations. Findings from this study indicate that a further investigation is needed for each practice and outcome variation. In addition, there is a need to minimize practice variations through the expansion of clinical practice guidelines and advocate their use, single-out some expensive but cost-effectiveness interventions from capitation payment with additional payment schedule. Using routine report and monitoring practice variations among peers such as caesarean section in Thailand tends to be practical and effective.

Keywords: provider payment, universal coverage, health insurance, inequity in health, efficiency

JEL Classifications: J18

Working Paper Series