Health Finance

Healthcare payment reform in the GCC – Abu Dhabi and Dubai’s move to DRGs

By Homer Warner, Jr. PhD
and James Vertrees PhD

There are good reasons why more than 30 of the world’s wealthiest countries have moved away from a fee-forservice healthcare payment or reimbursement to some form of prospective payment DRGtype scheme. Diagnosis Related Groups, or DRGs, are based on a simple concept of averages. With DRGs, government or insurance payers simply calculate the average cost of treating patients with similar characteristics (DRGs identify patients with similar clinical profiles and resource requirements for treatment) and pay the provider (the hospital in most cases) the average amount, regardless of what it costs to treat an individual patient. This payment method has proven effective for driving efficiency into inpatient healthcare delivery and has provided a mechanism whereby providers share some of the financial risk with the payers. In 2007, Abu Dhabi’s government hospitals made the decision to move to DRGs and this year, Dubai announced a similar move - at least for hospital funding.

The United States was the first to use DRGs for payment. When the US introduced DRGs in 1983 as their way of reimbursing its Medicare (elderly) patients, they saw healthcare annual inflation drop from 14% to 3%, and hospital length of stay drop by 20%. The number of annual hospital admissions decreased by 35 million patients and the US realized annual healthcare savings of $18 billion (according to a Rand Corporation study done in the 1980s). In addition to providing efficiency incentives for healthcare providers, DRGs relate funding to the mix of patient cases a hospital actually treats (its casemix) as well as providing comparative information, which supports internal management efforts to improve the quality of patient care.

Patients treated at hospitals differ based on severity of illness, prognosis, treatment difficulty, need for intervention and resource intensity. Even though individual patients are unique, there are similarities among groups of patients. DRGs provide a method for classifying patients, both in terms of their clinical similarity and the effort and cost required to provide them with treatment. Analysis of the patients by DRG helps us find differences (variance) in outcomes, quality, and costs of care. The number of DRG groups is also important. If there are too many DRG groups, there will be some with insufficient patient volume to provide statistical validity; if there are too few DRGs, the groups become too broadly defined, and important differences between the patients are concealed. Hence, the number of groups in most DRG systems is usually between 500 and 1000.

Because the average cost of treating a patient in a hospital can not be directly compared with that of another hospital, DRGs use an index, called the Relative Weight, to measure the cost appropriate for treating patients in a DRG group relative to the average patient cost. If one hospital treats sicker patients than another hospital, one would expect its average cost per patient to be higher. The Case Mix Index (CMI), which is the average Relative Weight for all of a hospital’s patient volume, is a measure of how sick (on average) the patients are in that hospital. By adjusting the hospital’s average cost by patient for its mix of patients (that is, by dividing this by the hospital’s CMI), we get a value that can be used to compare one hospital’s efficiency with that of another.


For example, when comparing the average cost of 17 hospitals as shown in the chart below, Hospital 14 appears to have the highest overall average cost (see the blue line). However, when the average cost is adjusted for the type of patients actually treated (its casemix) then the picture changes as shown in the red line and the average costs for Hospital 14 are more in line with the overall average, once adjusted for the hospital’s casemix.

The CMI may be useful for determining appropriate payment or reimbursement, but it is also useful for quality improvement (QI) and utilization review (UR). In hospitals where DRGs are used, outlier cases are automatically targeted for QI/UR and over time have helped reduce practice variance among physicians. They provide hospitals a means for statistical process control.

DRGs utilize routinely collected information (i.e., diagnoses and procedure codes, age, sex, length of stay, discharge disposition, and birth weight for infants). In countries where DRGs are used for payment, a standardized coded minimum data set is collected from all hospitals and there is increased emphasis on accurate, consistent, and complete medical record documentation and coding. DRG assignment and reimbursement are directly dependent on coding quality.

The coding of diagnoses and procedures after a patient is discharged from the hospital is usually done by trained medical records coders who are familiar with standard disease classifications like ICD-9-CM or ICD-10 (as defined by the World Health Organization). In the US, Canada and Australia, the typical ratio of coders to number of hospital beds is 1:50 (however, this can vary based on what the responsibilities of the coders include).

By comparison, many hospitals in the Gulf region have too few coders to keep up with the work load and lack coding software to help increase accuracy and consistency. As GCC hospitals begin to prepare for the adoption of DRGs (in some form), they will need to train their physicians to improve their chart documentation (e.g., include complications and comorbidities along with the primary or principal diagnosis, relevant secondary diagnoses and all significant procedures) and invest in coder education, more coders and coding tools. With the collection of accurate and complete coded patient data, DRGs can provide a view into what is going on within hospitals and between hospitals and physicians.

Of course, DRGs, like all things in this life, are not perfect. For example, under DRGs, incentives for efficient behavior extend to a hospital discharge but not beyond. This means that, at least in theory, a DRG-based payment system gives hospitals a financial incentive to discharge patients “quicker and sicker”. Despite the fact that quality of care has not measurably declined in any of the many countries that use DRGs for payment, this is and should be a concern. The most important factor offsetting this incentive is the ethics of the health care professionals involved in the patient’s care. However, it is also important to put strong quality assurance measures in place along with the DRGs. Indeed, these measures are needed in any payment system (even fee-forservice) as higher cost does not always represent better quality.

In addition, it is important that any patient who feels that he/she is being discharged before they are ready to go home have an avenue to appeal the discharge. Hospitals in the US are required to inform patients of their rights and to provide an independent review of disputed cases where early discharge is problematic – even if the problem is related to the resources available at the patients’ home. It is also important that the system uses modern severity adjusted DRGs for the payment system, in order to recognize that some patients are more severely ill and are likely to require more resources for treatment. Several of these severity adjusted systems are currently available in the market.

It is also true that DRGs do not give providers an incentive to reduce the number of admissions. Again, while increases in admissions have not been a serious problem in the countries that use DRGs, it is important to put in place utilization review measures that determine admission appropriateness along with the DRGs. Indeed, it is useful to note that these measures are even more important in the absence of DRGs.

In a similar vein, DRGs are intended for inpatient care. DRG-based payment systems do not reward, nor do they penalize, preventative care. Prevention efforts usually include anti-smoking, exercise, weight control, pharmaceutical compliance, and childhood vaccination, programs, and these are often outpatient services. DRGs are not inconsistent with prevention efforts, but they do not facilitate these efforts either.

Finally, it is important to recognize that classifying psychiatric care and inpatient care for drug and alcohol abuse are difficult. For this reason, these types of inpatient care are issues for the DRGs. The problem is that two doctors may provide different psychiatric diagnoses for the same patient and, given the same diagnosis, the same two doctors may treat the problem differently. For these reasons, inpatient psychiatric care and inpatient care for drug and alcohol abuse are often used in limited ways or excluded from DRG-based financing systems.

Thus, it is very important to recognize that DRGs, like all real world payment systems, have limitations. The real question is whether a DRGbased system will be better for providers, payers and patients as compared to the system currently in place. As Dr. Robert Fetter, the inventor of DRGs, always said; “One should never let the perfect become the enemy of the good”. Realizing that DRGs are not perfect, both Abu Dhabi and Dubai Health Authorities have decided that the tradeoffs favor the use of DRGs and expect to see improved inpatient healthcare efficiency, cost reductions, and higher quality data for comparing healthcare delivery among providers.

The authors:
Homer Warner, Jr. PhD is the GCC business development manager for 3M Health Information Systems and James Vertrees, PhD is a Senior Health Economist for 3M Health Information Systems.

ate of upload: 31st March 2009

                                               Copyright © 2009 All Rights Reserved.