Payments to healthcare providers entail a set of economic incentives that influence provider behavior and decision-making [1, 2]. Israel adopted activity-based payments to replace per diems (PDs) and created codes for 30 common procedures as early as the 1990s [3]. The main objective of the change was to shorten waiting times for expensive procedures involving brief hospital stays, for which the PD payment was insufficient so that hospitals were discouraged from performing them [4]. Due to data and policy constraints, Israel chose procedure-related groups (PRGs) rather than Diagnosis Related Groups (DRGs) as the basis for measuring activity. PRGs differ from DRGs in that they are defined based on type of treatment (surgical procedure) rather than diagnosis, and they are not adjusted for case-mix or disease severity [5].
In the past two decades, many OECD countries have shifted to hospital payments based on activity and adopted diagnosis-related groups (DRGs) as payment units but, unlike the Israeli case, their main objectives were to increase efficiency and transparency [6]. DRGs are still being adopted by mid-income countries [7]. In 2002, continuing the move towards activity-based payments, the Israeli Ministry of Health (MoH) created PRG codes for more procedures, in the same timing DRGs were introduced in some OECD countries such as Estonia, Germany and the Netherlands [6]. Since 2010, the MoH has further expanded the application of PRGs to several clinical specialties, in three main waves:
-
1.
Wave 1: 2010–2012, trauma in orthopedics
-
2.
Wave 2: 2013–2014, urology, general surgery, ophthalmology, head and neck surgery
-
3.
Wave 3: 2015, orthopedics and MRI
The objectives of the 2010–2015 “PRG reform” mainly concerned transparency and a fair distribution of funds. The specific objectives were to refine the unit of payment and establish consistent costing and pricing mechanisms in order to reduce cost-price gaps, improve MoH ability to set policy and priorities, influence the supply of hospital services by adjusting prices, and conduct supervision and control [5]. Furthermore, PRG payments were expected to change the incentives for hospitals. If PD payments create incentives for longer stays, PRGs create incentives to perform more procedures and shorten the length of stay (LoS), to minimize operating costs and maximize profits.
Many studies have evaluated the impact of DRG-based payments in high- and middle-income countries on volume of activity, LoS, and quality of care [8,9,10]. In Israel, Shmueli and colleagues [11] examined the effects of the early introduction of PRG payments for five major procedures, one year after implementation in 1990. They found that the volume of activity increased for two procedures, remained unchanged for two others, and decreased for the last one. Regarding LoS, there was a modest decrease in three procedures and a significant decrease in the other two. A later study evaluated the effect of incorporating the time interval between hospitalization and treatment (of hip fractures) in the PRG tariff (maximum fees are paid for patients operated within 48 h, for those operated later, payments are significantly lower); it found that the LoS decreased following this change in payment method [12].
Since then, no study has evaluated the effects of the later adoption of PRGs on hospital activity. The effects of the 2010-reform thus remain largely unknown, preventing evidence-informed discussion of its benefits and challenges. The current study adds to the previous literature both by analyzing the changes that have occurred since then, and extending it, by examining all the hospital data and including all the activities performed at the ward level.
Background on the Israeli case and the hospital market
Since 1995, Israel has had a national health insurance (NHI): four competing, non-profit health plans (HPs) are responsible for providing and managing a broad benefits package determined by the government. The HPs provide care in the community and purchase hospital services for their members.
Of the 44 general hospitals in Israel, 35 are non-profit and owned by the Ministry of Health (MoH), the municipalities, the Clalit HP or NGOs. These are considered “public hospitals.” The other nine are smaller, for-profit hospitals, and operate 3% of the beds. The main source of income of Israeli public hospitals is the sale of services to HPs and the National Insurance Institute (NII) (see left-hand column in Fig. 1). Hospital reimbursement rates are determined by a joint MoH and Ministry of Finance (MoF) pricing committee, stipulated in the “Price List for Ambulatory and Inpatient Services.” This maximum list-price (tariff) also determines the type of payment, which can be PD; per activity (PRG); or fee-for-service (FFS) (see right-hand column in Fig. 1). There are currently 24 PD rates according to ward type and length of stay (the tariff of the first three days is higher than the tariff of the subsequent days), about 320 PRG codes, and more than 1600 ambulatory service codes. In 2015, 25% of the gross revenue of hospitals was for inpatient care paid as PRGs, 37% for inpatient care paid as PDs, 21% for ambulatory care paid as FFS or PRGs, 8% for births paid as PRGs, 6% for emergency care paid as FFS, and 3% from other sources such as the Ministry of Defense or the military [13].
The sale of services covers hospital marginal costs and some fixed costs such as physician salaries. Public hospitals also receive “prospective subsidies” in the form of global budgets from the MoH to cover part of the other fixed costs such as infrastructure and equipment. Furthermore, the government provides “retrospective subsidies” for public hospitals with a financial deficit at the end of each year. Both subsidies are negotiated with both the MoH and MoF. Overall, hospitals received about NIS 1500 million, which roughly represents 12% of their income from government subsidies (yellow box in Fig. 1) [13].
Israeli public hospitals are subject to two major income constraints. The first, put in place in 2005, is a cap mechanism; the MoH sets annual caps on hospital revenues from each HP to each hospital (see vertical arrows at right-hand side of Fig. 1). In recent years, caps have been set as a floor (lower cap bound) and a ceiling (upper cap bound), and are updated every three years. The floor is a minimum payment amount, set in 2016 as 93% of the previous year’s expenditure for each HP to each hospital. If an HP consumes services that, at “list prices”, would have an aggregate cost of less than the lower cap, the HP pays 93% of the previous year’s expenditure to the hospital in any case. The ceiling is a maximum payment amount and when an HP spends more than this threshold, it pays only a percentage (less than 100%) of the full price [14]. Towards the end of the financial year, once the upper bound of the cap is reached, there is an incentive for HP to refer patients, when possible (e.g., for elective procedures) to the hospital as they do not pay the full price for these services. In 2016, the hospitals’ net income was 15% lower than the potential gross income due to discounts related to the cap mechanism [13]. The second constraint is a negotiated alternative reimbursement contract between an HP and a hospital that may supplant the official cap, with such contracts entailing discounts that vary across HPs and hospitals [4]. In 2015, individual discounts represented 4% of the hospitals’ gross income [13].
Acute hospital care in Israel has a high rate of overcrowding, one of the highest among OECD countries. Compared with the OECD average, Israeli hospitals function with half the rates of acute-care beds and nurses per population. In 2017, the average length of stay (ALoS) in Israeli hospitals was 4 days, one of the shortest, and occupancy rates of acute-care beds was one of the highest among OECD countries, reaching almost full capacity, 93%. Nonetheless, the number of discharges per 100,000 population in Israel is almost the same as the OECD average [15, 16].
Objectives
Our objective was to examine changes in the volume of activity, measured by the number of discharges and ALoS, in hospitals following the PRG reform. The focus was on changes on the macro/system level, aiming at draw generalizable conclusions about the payment-policy change rather than an examination of the impact on specific procedures or hospitals. Since PRG codes were created in waves by clinical area, we hypothesized increasing volumes and decreasing ALoS in the clinical areas for which PRG codes were created. Our analysis focuses on hospital wards as a proxy for such clinical areas.
Economic theory suggests that hospitals react to economic incentives derived from payment mechanisms [17]. Peleg and colleagues [12] show that immediately after the adoption of the refined PRG codes, Israeli hospitals reacted by costing the wait between injury and surgery for timing of hip fracture procedures. Based on international experience [9], it is quite plausible that this was a reaction to the PRG reform. However, in contrast to other OECD countries, where DRGs were adopted to improve efficiency, Israeli hospitals operated with relatively limited resources even before the adoption of PRGs in 2010, potentially limiting the hospitals’ capacity for increased activity. The analysis of the effect of the PRG reform on hospital volume and ALoS, in so different an environment, thus provides an interesting case.