Copayment and recommended strategies to mitigate its impacts on access to emergency medical services under universal health coverage: a case study from Thailand

Over the last 3 decades, EMS systems in Thailand have gradually evolved in response to the emergency health needs. The EMS systems include pre-hospital ambulance services, medical emergency dispatch centers, ED services and inpatient definitive care organized mainly by public providers in rural and urban settings and to a certain extent by private providers in major cities.

Since 2001, UCS, SSS and CSMBS have mainly financed hospital emergency services including ED and inpatient definitive care. The payment methods and payment rates differ among the 3 schemes with some overlapping as follows.

Under the UCS, all the contracted hospitals, both public and private, are paid for EMS provided to a non-registered UCS member by fee for service (FFS) with a ceiling of 700 Baht for ED visits and DRG with global budget for inpatient care. FFS with ceiling was adopted to pay private hospitals outside the contract for both ED and inpatient care. Under the SSS, public hospitals are paid according to the billing for ED visits and the first 72 h of admission. For private non-registered hospitals, both under and outside the contract, the patients have to pay upfront and reimburse from the central fund by FFS with ceiling for both ED visits and the first 72 h of admission. After 72 h of admission, the registered hospital is fully responsible for the costs of service provided.

Under CSMBS payment methods, all public hospitals are reimbursed on a fee schedule for ED visits and on the DRG basis for inpatient care plus boarding and medical supply based on a predefined list. Prior to the policy initiative, the reimbursement for inpatient care to private hospitals was capped at 4,000 Baht (133 United States Dollar (USD)) per admission. There was no payment for ED service to private hospitals. Notably, a CSMBS beneficiary is not required to register to any hospitals, as opposed to those of SSS and UCS. Should there be a need for inter-hospital transfer after initial admission to a hospital for an emergency condition they are more likely to face difficulty in seeking another hospital for the transfer.

Without an exclusive dataset for ED services, an unofficial estimate of total ED visits to public hospital was reported to be 13 million annually [15]. Based on this figure and a reported figure of 32 % admission rate from ED [16], the annual workload of inpatient emergency cases for the whole country could be at least 416,000 per year. Despite the fact that private providers play a significant role in provision of emergency services in the country, there is no data on the service provision and expenditure [17].

The unified EMS policy aims to close the gap of access to EMS under the three health insurance schemes. The policy interventions include mobilizing private hospitals outside the previous contracts; unification of payment methods and payment rates for purchase of EMS from these private hospitals; public policy communication to inform the beneficiaries on the right of access without copayment; and monitoring of the progress for ongoing adjustment. The National Health Security Office (NHSO), in charge of UCS, is assigned to be the responsible agency in executing this policy. The unified purchasing model is characterized as single payment methods and single payment rates for ED visits and hospitalization to the outside contracted private hospitals. For ED visits, the payment is based on the fee schedule of the existing CSMBS protocol plus extra items beyond the fee schedule, but does not exceed 1,000 Baht. For hospitalization, DRG is a case-based payment at 10,500 Baht (approximately 350 USD) per adjusted RW which is used to reimburse the private hospitals. The private hospitals receive extra payments for high cost medications and medical instruments (e.g., prosthesis, intravascular stents). Financial audit of every claim is mandatory prior to the reimbursement. This includes checking severity classification, DRG related data, health insurance entitlement, patient identification (ID) code, hospital ID code, and date/time of hospital visit and discharge. NHSO also acts as the Clearing House to process reimbursement claim, patient complaints, and to coordinate, in collaboration with Ministry of Public Health (MOPH), inter-hospital transfer of patients. An administrative dataset was exclusively set up to facilitate the financial claim processes using online transactions.

As part of the policy implementation, the government made use of mass media and out-door media for public communication campaign to raise awareness of free access to all public and private hospitals nation-wide in case of a need for EMS despite the fact that supply-side interventions are limited to 255 outside contracted private hospitals (78.2 % of the total number of private hospitals). In addition to the unified payment mechanism, the MOPH held several official meetings with representatives of the private hospitals to press them to participate citing the mandate stipulated in The Sanatorium Act B.E. 2541. The law requires private hospitals to provide emergency care for any patients until the emergency condition has been stabilized without mentioning any hospital payments and definitions of the conditions. The combined supply-side and demand-side interventions of the programs mentioned are expected to work together in promoting better access to emergency services. It is noteworthy that the majority of the outside contracted private hospitals previously chose to focus on the high-end well-paid market of healthcare. Setting the payment rate without adequate participation of these private hospitals hence posted a challenge in their compliance to the service provision. The lack of any regulatory capacity to compel the private hospitals not to charge copayments above the EMS payments posted another challenge of copayment.

In practice, the program involved a large body of representatives mainly from the outside contracted private hospitals, 3 departmental units of the MOPH, and the 3 schemes. In every high-level decision-making forum, the Prime Minister chaired the meetings herself indicating a strong policy commitment. Monthly reports of the progress in policy implementation in terms of outputs and public responses (directly from the complaint channel and indirectly from mass media) were key inputs to the forums. Multiple stakeholders with multiple perspectives were brought into the policy arena along with the ambitious purpose of achieving more unified healthcare financing mechanism (after a decade of stagnation on this policy track) [8] and access to EMS expanded at the same time to support this policy initiative.

Analysis of the EMCO dataset reveals 43,588 patient claim visits to 225 private hospitals (88.2 % of the total eligible 255 private hospitals) during the first 21 months of the program (April 2012 - December 2013). The majority of claims (91 %) were for inpatient care. Table 3 shows CSMBS beneficiaries constituted the majority of total inpatients (59.8 %) then UCS beneficiaries came second (34.1 %) and SSS beneficiaries were the smallest proportion (6.1 %). This is discordant to the percentage of total number of beneficiaries for each group. The number of inpatient claims for each month varied from 1,232 to 2,750 cases with a fluctuating trend and peaked at the 9th month. While percentage of access of CSMBS beneficiaries started at the lowest in 1st month then increased 1.8 times later on, those of UCS and SSS beneficiaries showed the opposite trends. Regarding severity among the three groups, CSMBS beneficiaries were hospitalized with the lowest percent (36 %) of critical conditions as compared to those of the others (Table 3).

The majority of patients reportedly visited with urgent conditions (22,569 cases or 53 %) shown in Table 4. Overall, those with critical conditions encountered a fatality of 11.7 % which is 10 times more than those with other conditions. However, the trends of fatality for each severity group were not stable. The last five months observed a sharp drop in fatality of those with critical conditions, roughly a third of those in earlier months. This happened in parallel with a sharp increase in the figures for those with urgent or non-urgent conditions. These findings indicated misclassification in severity of the claim records. Nonetheless, CMI seemed to follow a steadier declining trend. The percentage of cases for each 10th International Statistical Classification of Diseases and Related Health Problems (ICD-10) category with top three highest figures respectively was 21 % for the respiratory system, 15.2 % for digestive system, and 14.6 % for certain infectious and parasitic diseases.

Hospitals participating in the program varied in size from 10 to 500 beds with a median of 150 beds (top quartile 230 beds and bottom quartile 100 beds). Over the period, claims from small hospitals (under 200 beds, Fig. 1) dominated the whole picture. While the number of claims from big hospitals (over 300 beds) dropped continually after the 4th month, those from other hospitals seemed to follow an inverted U shape with 2 consecutive sharp drops at 10th month and the 16th month respectively. Hospital charge per case varied from 1,000 to 5.5 million baht (33.3 to 183,333.3 USD using an exchange rate of 1 USD for 30 baht) with a median of 28,018 Baht (934 USD). Average charge positively correlated with hospital size (upper panel, Fig. 2) which was similar to the correlation between adjRW and hospital size (lower panel, Fig. 2). In terms of payment, hospitals were reimbursed at a minimum of 0.0002 to a maximum of 9.9 times the total amount of charge with a median of 33.5 % of the total amount. At the top quartile, hospitals were reimbursed at 56 % of the total amount and 19.8 % at the bottom quartile. Faced with a perceived under payment, hospitals might react by upward adjustment of the charge.

On one hand, average monthly hospital charge showed a fluctuating trend over the 21 months (Fig. 3), whereas monthly CMI seemed to follow a slight downward trend (Table 4). On the other hand, paid-charge ratio followed a sharp downward trend during the same period, (Fig. 3) together with a drastic drop of the approved claim rate from 82.5 to 29.6 % during the period (overall claim approval was 59.3 %) (Table 5). These seemingly contradictory pieces of evidence from each side might indicate upward adjustment of hospital charge per case or per service item given fixed reimbursement rates. It might also indicate changing mix of hospital size with a tendency of big hospitals (over 300 beds) withdrawing from the program quite early (Fig. 1). These observations are in accordance with further analysis as follows. Over the 21 months, percentage of monthly claims being rejected or pending was steadily increasing from the first month onward (Table 5). Using multiple linear regression analysis (Table 6), it was found that the total amount of hospital charge was significantly associated with the following factors (in order of magnitude according to standardized beta coefficient): adjRW, paid charge ratio (ratio between total amount of reimbursement and total amount of hospital charge), age, hospital size, hospital location, month of hospital visits, and health insurance status except for sex and severity. This means the amount of hospital charge increased according to the value of adjRW. Big hospitals charge higher price than small hospitals (Fig. 2). Hospitals in Bangkok charge more than hospitals outside Bangkok although consumer price index for Bangkok was not the highest in the country during 2012 and 2013 [21]. The amount of charge followed an increasing trend from the beginning to the end of the 21-month period. It increased against paid-charge ratio and increased with age of the patient. The amount decreased when tracking from that for CSMBS beneficiaries to those for SSS and UCS respectively.

Up to this point, the interactions between the hospitals and the Clearing House might be better understood taking the viewpoint of complex adaptive systems (CAS) [22]. According to this viewpoint, healthcare system is considered a system which consists of elements interacting in freedom and with the ability to respond to stimuli in many different and fundamentally unpredictable ways. Adopting the CAS perspective, it follows that the hospitals play out the game by increasing charges or withdrawing (Fig. 1) based on perceived under payment and the payers (through the Clearing House) counteract by claim rejection or adherence to the payment rates. Additionally, the hospitals could also employ copayment as a strategy to cope with perceived under payment given the absence of laws or regulations precluding copayment. The next section is going to deal with this notion.

With the perceived under payment, hospitals might opt for copayment in addition to the upward adjustment of charges or withdrawal. Based on the paid-charge ratios, it could be inferred that the hospitals might request for a substantial copayment from roughly 40 to 80 % of the total charge. This notion is supported by evidence from the consumer surveys. As part of the feedback to the policy implementation, the Clearing House ran a series of telephone surveys in 3 consecutive periods (1st–3rd months, 7th–9th months and 12th–15th month), which showed the prevalence of copayment varied from 53 to 95 % as shown in Table 7. Results from our telephone survey (N = 128) for the first month of the program found the prevalence (%) of copayment varied from 80 % for CSMBS beneficiaries to 50 % for UCS and 30 % for SSS beneficiaries. The prevalence of copayment requirements reflects the fact that private hospitals perceived the reimbursement rate set by the Clearing House as under compensation to the cost of services.

The discrepancy in the amount reimbursed and the total charge clearly demonstrates a conflict between pricing policy of the hospitals and the adopted payment policy. To support the pricing policy, the hospitals argued that they set prices to reflect full cost recovery and a surplus of under 15 % per annum (Table 8). Based on this argument, we explored our previous report on pricing of the hospitals using EMCO dataset for the first 3 month period focusing on patients in critical or urgent conditions (N = 1,257 records). We found different prices per adjRW for the services among 3 categories of hospitals: security-market registered hospitals, general private hospitals, and not-for-profit private hospitals. On average security-market registered hospitals are bigger than the others (174 vs 81 beds per hospitals). The median price varied as follows: 47,246 Baht, 38,565 Baht and 35,727 Baht, respectively (p-value = 0.0001). Further analysis taking account expected differences in fixed and operating cost between hospitals in Bangkok and regional provinces, we found no difference in price among security-market registered hospitals operating in Bangkok or regional provinces. To the contrary, the differences were demonstrated for the other two categories i.e., higher price for hospitals operating in Bangkok and lower price for hospitals in regional provinces. These findings vividly indicate different pricing policies among different categories of the hospitals. It should be noted that security-market registered hospitals charged the highest price despite the fact that they are in a better position to access loans at the lowest interest rates compared to the others. Finally, in the same report, our team conducted a head-to-head comparison of itemized prices (medications and medical supply) on 80 selected patients of the private hospitals to those prices of a tertiary care public hospital to shed light on rational of the pricing. We found that medication prices differed from 1.15 to 399 times and 1.55 to 4.4 times for medical supply. By reviewing medical records and hospital bills, we also found seemingly unnecessary charge items on the hospital bills. For instance, the items included a screening test for prostate cancer in a patient with pneumonia, bone fixation in a patient with hypertension, and blood transfusion without proper medical indications. Results of the head-to-head comparison and the review of medical records cum hospital bills suggested inappropriate pricing of some private hospitals. After viewing these findings, reaction from representatives of the hospitals turned out quite predictable. They argued that private hospitals have to fully rely on themselves from the first to the last dollar in capital investment and daily operation, which is not the case for public hospitals. As a result, they argued comparisons are not appropriate. With regard to the findings from the review of medical records and hospital bills, they did not give any opinion.

In-depth interview provided further evidence related to the copayment level. Private hospitals cited unclear operational definitions of emergency conditions as a reason to impose copayment to deter perceived abuse of the services by patients. This perception emerges from the hospitals report of encountering patients with non-urgent conditions requesting free access to ED and inpatient care. This notion sounds logical given the large proportion (90 %) of patients accessing ED by informal means [15]. This renders the prehospital care triage systems (with standard criteria of severity classification) irrelevant. Next, the hospitals also reported patients bypassing closest public hospitals to their facilities. This finding corresponded to the finding from our telephone surveys (N = 128) revealing the most common reasons for bypassing closest public or private hospitals were due to a perception of inferior quality of care. Finally, the private hospitals vindicate copayment by attributing to the lack of criteria to justify inter-hospital transfer after the first 72 h of admission while the patients demand further stay.

Our telephone survey on the 128 patient samples revealed over 90 % of them were ignorant of the channel for making complaint to the Clearing House. During the early months of the program, analysis of complaint calls from the patients or relatives to the Clearing House identified 1,165 calls asking for clarification of eligible emergency conditions; 633 calls of eligible hospitals for access; and 409 calls of the conditions for copayment. These findings indicate inadequate public policy communication to the users.

The high degree of freedom in price setting by the hospitals raises an issue of hospital regulatory mechanism. The Sanatorium Act B.E. 2541 (section 32) requires hospitals to disclose prices of medical and hospital services which hospitals have to comply with accordingly. To enforce the laws, the Health Facility Committee has been set up and chaired by the Permanent Secretary of the MOPH. Its voting members consist of the Heads of 4 departments of the MOPH, a representative of the Council of State and of the Office of Consumer Protection. In addition, the committee includes 3 health practitioners, a representative of each of the health professional councils and technical advisors which have to be representatives from the private hospitals for at least 1 person and no more than 3. The Committee has been tasked with issuing regulations, granting permission for establishing and operating a hospital, promoting quality improvement and processing appeals. Since the law has been in place for almost two decades, there have not been any information systems to keep track of the hospital behaviors. Alarmingly, informal interview found a key informant, (a senior health officer with direct experience in working on the committee) who maintained that he had never witnessed a single case of hospitals being seriously dealt with for misconduct. This notion is in keeping with findings from the review of the annual reports in 2006–2014 of the department overseeing the Sanatorium Act, which found no evidence pertinent to the law enforcement [23].

Turning to the payment rate issue, in-depth interview with high-ranking officers of the Clearing House and the health insurance schemes revealed the rationale behind the DRG-based payment rate. We find the rate was being set without concrete evidence about the cost of hospital services except for a concern of future financial burden to the UCS if a higher rate than the current one would have been adopted. However, in effect, a higher rate of DRG-based payment to private hospitals has been adopted for other previous programs under the UCS for medical services for premature newborns, patients with severe head injury, cardiac catherization etcetera.

Finally, in-depth interview of the administrators of the 3 schemes revealed no concrete evidence of any progress in the attempts to make changes in rules and regulations pertinent to the program requirement i.e. harmonizing the payment mechanism and payment rates to facilitate access to EMS of the hospitals. Hence, the existing payment made possible using the financial resources and personnel of the UCS assigned to take major responsibility in the program execution through the Clearing House. Major obstacles cited by the administrators of SSS and CSMBS are the concern of potential impacts of the expected amendment of rules and regulations on all other benefit packages for non-emergency conditions.

Using a mixed methods, this report is able to shed light on the existence of copayment despite the public financing of access to the hospital provision of EMS during the 21 months of the program implementation. The copayment existed in parallel with inequitable access to EMS among the beneficiaries of the 3 schemes as shown in Table 3. A previous survey of 3,504 ED users in public hospitals in the 4 regions of the country revealed different percentage of access among the beneficiaries in contrast to the figures. in Table 3 i.e., 10.2 % for CSMBS, 6.0 % for SSS and 68.5 % for UCS [24] Similar figures from an earlier ED survey of 6,440 patients was also documented [12]. The copayment also coexisted with inequitable health outcomes as reported in a previous study based on analysis of the EMCO dataset [15]. Evidence from developed countries on the effects of copayment on access to hospital ED or other forms of healthcare, nonetheless, has been inconclusive as mentioned in the introduction. A recent report using difference-in-difference analysis of data over a 10-year period did not demonstrate a reduction in ED use following the enforcement of copayment in the U.S. [20] The authors ascribe this negative finding to several factors. First, difficulty in collecting the copayment out of fear of violating the provisions of the Emergency Medical Treatment & Labor (EMTALA) Act Law, which requires hospitals to provide appropriate medical screening to persons seeking medical care through an ED, and that hospitals treat and stabilize anyone with an emergency medical condition. Second, ED health care providers may have difficulty determining whether a visit was due to a medical emergency or was non-urgent. There has been no consensus on what constitutes a non-urgent visit, which helps to account for the wide variation in estimates for non-urgent ED visits (range, 8–60 % of all ED visits). Third, ED staff are required to give the beneficiary the name of an accessible Medicaid health care provider, which, given the national shortage of Medicaid health care providers, may be difficult for many EDs.

In similar to several countries in Organisation for Economic Co-operation and Development (OECD) with UHC, under the policy directive Thai patients in urgent conditions are allowed to directly seek hospital ED without copayment [25]. However, the policy put emphasis on seeking ED closest to the place of emergency events. To the contrary, our findings documented hospitals concerned about bypassing issue and used this issue to justify copayment (Table 8). Viewing from the demand side perspective, our telephone surveys found perception of inferior quality of care as main reason for bypassing. In the Netherlands, Valk et al. found expectation of diagnostic tests not available at general practice clinics as the main reason for patients bypassing to hospital ED [26]. The next common reason was patient being already under specialist care at the hospital. More or less, these findings could be considered in the same vein as our report of the patients’ perception of inferior quality of care. Perhaps, these reasons may influence some countries (such as England and Denmark) with strong gatekeeping mechanisms in place to introduce new initiatives designed to facilitate access to specialists [25].

The declining percentage of patients with critical conditions over the 21-month period shown in Table 4 may reflect a demand for after-hour specialist care. Schoen et al. reported difficulty in getting care nights, weekends, or holidays was of significant concern among adults in all five OECD countries (Canadian and U.S. adults especially the elderly were also more likely to have gone to hospital ED for care that their regular source could have provided if available [27]. This finding is in keeping with our results shown in Table 3 that CSMBS beneficiaries not only constituted the highest percent of ED users but also were the oldest among all groups of beneficiaries. Taking together the findings of OECD countries and ours, ED could be deemed a sensitive indicator for how healthcare systems are responding to patients’ needs. And in response to after-hour healthcare needs, those OECD countries came up with different initiatives such as after-hours general practitioner clinics co-located with ED in Australia, a Primary Health Care Transition Fund to encourage models of round-the-clock, team-based care in Canada, and twenty-four-hour nurse-staffed hotline and NHS Walk-In Centres in the United Kingdom [27]. Similarly, many public hospitals in Thailand currently provide after-hour clinics from 5 to 8 pm on weekdays and 9–12 am on Saturday. Nonetheless, there has not been any systematic assessment as to the impacts of these after-hour clinics in meeting the demand using ED as an indicator.

These difficulties confronting the U.S. hospitals (mainly private) are, in turn, instructive to improvement of the provisions and enforcement of the Sanatorium Act in Thailand. First the Sanatorium Act should include provisions with more specific details requiring hospitals to provide screening, stabilizing and transfer for the patients similar to those of EMTALA. Provisions for penalty should also be clear and substantial enough like that of EMTALA. Apart from the provisions; the effectiveness of enforcing EMTALA lies on the so-called “Medicare death penalty” i.e. termination of Medicare contract to hospitals based on result of investigation of potential law violations by the Centers for Medicare and Medicaid Services (CMS) [28]. The death penalty is meaningful to the United States hospitals because of arguably substantial financial contribution from CMS to the hospitals. This is not quite the case in Thai context in which major revenues for many private hospitals come from private insurance and OOP. In addition, with diverse payment rates and methods among the 3 schemes functioning independently, it is difficult to impose meaningful financial sanctions to the private hospitals. In this regard, harmonization of payment methods and rates of the 3 schemes should take into account strengthening the bargaining power of public healthcare financing against the private hospitals. Finally, enforcement of the Sanatorium Act would need to put in place an investigative mechanism like that of CMS. This involves impartial status of the investigators and sufficient financial support for the investigation.

Concerning well-informed decision of the users, the findings on limited knowledge of the users indicate inadequate policy communication based mainly on mass media and out-door media. This could be addressed through regulatory measure mandating the hospitals to make formal notification of the right of access to every patient visiting ED as is the case under EMTALA. In many European countries, hospitals are required to bear the burden of proof in case of liability for malpractice and errors in providing information [29]. This kind of legal provision could enhance the sensitivity of the hospitals to users’ need for relevant information.

Review of national efforts among BRICS countries (Brazil, Russia, India, China, and South Africa) reveals substantial contribution of private providers in healthcare provision [30]. In fact, this situation is not unique to BRICS countries but is similar to those in many other low-middle income countries including Thailand. This undeniable role of the private sector poses a specific challenge in the public sector i.e., stewarding mixed private and public health systems. Our report suggests intermingled role of stewardship (the role of regulator) and financing (the role of the three major schemes) in promoting more access to EMS in Thailand - a country with over a decade of experience in sustaining UHC. Although we are able to shed light on provision of EMS among a selected group of private hospitals not previously included in the purchasing models of the three schemes, regrettably so far there has not been any evidence on the provision of EMS by contracted private hospitals under SSS and UCS. This indicates room for improvement of relevant information systems to keep track of the service provision.

Considering a need for strategic purchasing from powerful private sector in pursuit of UHC, the present report indicates a number of inter-related key specific issues for further studies: capacity building for healthcare financing agencies in negotiating with private providers based on rational and evidence-based price setting; and design of technically strong and high commitment of regulatory mechanism to set up effective legal framework to ensure sufficient access and fair pricing. Both these issues should be conceived along with the development of sufficient information systems to keep track of private providers in terms of service provision (access and quality) and pricing. Finally, further studies should focus on assessment of implementation of the suggested legal measures to enhance providers’ sensitivity to users’ need for relevant information at the point of care. Finally, systematic assessment on the impacts of the provision of after-hour clinics by public hospitals should be conducted taking account ED as an indicator.

We chose to focus on inpatient dataset hence rendering inadequate exploration of ED visits. The results of user survey could be biased due to substantial number of non-respondents or failure to make contact. We could not explore methodological details of the telephone survey of the Clearing House, which might be subject to selection bias and information bias. Without access to minutes of the executive meetings during the policy implementation, we were not able to explore interactions among policymakers and the key stakeholders. In-depth interviews were conducted during the first few months of the program, hence the findings may not be up-to-date given the dynamicity of the policy context resulting in changing faces of policymakers during the 21-month period. Finally, administration of the EMCO dataset is not sophisticated enough, hence quite substantial relevant data are missing for some variables. This deficit in data quality might affect parameter estimation.