Doctors' Performance Measures


A doctor's performance is a key factor, in addition to prices, that patients will want to consider in selecting a doctor. In the RE·MEDI system, there is great emphasis on measuring performance and publishing these measures so that our members will be well-informed. We have developed several performance measures for both our Diagnosticians and treating doctors. For Diagnosticians, these include: the Referral Satisfaction Index (RSI), which measures the quality of referrals made by a Diagnostician; the Diagnostic Accuracy Index (DAI), which measures the accuracy and timeliness of a Diagnostician's diagnoses; and the Prognosis Rating Index (PRI), a measure to motivate Diagnosticians to issue similar prognosis ratings in similar situations.

For treating doctors, there are also three measures for assessing performance: the Patient Satisfaction Index (PSI) for measuring the extent to which doctors satisfy their patients; the Nurses Index, which identifies the doctors that nurses choose for their own treatment and for that of their family members; and the RE·MEDI Outcomes Index, which is an unbiased and risk-adjusted measure of how well a doctor's patients have recovered from a particular illness.

Performance of Diagnosticians

Referral Satisfaction Index. Whenever the Diagnostician prepares a treatment plan for a patient, he also recommends a doctor for the patient's treatment. This doctor and the RE·MEDI system's information about him are always included in the patient's copy of The Doctor Shopper. The requirement that the Diagnostician always recommend a treating doctor to the patient is important because, in the current healthcare delivery system, patients are not always well-served by the recommendations they receive from friends and doctors. The Referral Satisfaction Index (RSI) has been constructed to give the Diagnostician a strong incentive to recommend an exceptional doctor to the patient for her treatment. Because the other doctors listed in The Doctor Shopper are usually selected based on their proximity to the patient's home or place of work, the Diagnostician's recommendation, though perhaps less conveniently located, has the potential to represent exceptional value to the patient.

The Diagnostician's RSI is based on two components: the "distance" of recommended doctors from the Cutting Edge on the Value Chart, and patients' overall satisfaction with doctors recommended by the Diagnostician. The distance of a doctor from The Cutting Edge is a good measure to include in the RSI, with shorter distances being preferred, as these doctors represent better value. Also, the Community Edge is used. The Community Edge is the Cutting Edge on the Value Chart that includes all of the doctors offering the treatments the patient requires and who practice within the local healthcare community. If the recommended doctor is on the Edge, then we score this component of the RSI with a 1.0; if he is off the Edge, then the score is less than 1.0 and is lower the farther he is from the Edge. For points lying on the Edge itself, no preference is given for one point over another. Evaluating the tradeoff between price and performance is subjective in nature and therefore is more appropriately left to the individual healthcare consumer.

If price and performance were the only criteria for rating doctors, then those on the Edge would always be preferred to doctors off the Edge. However, criteria other than price and performance should be considered for inclusion in the RSI. Whether the doctor's office is convenient to the patient, and whether the personalities of the patient and doctor are compatible are just two examples. For the most part, these criteria are already reflected in the Patient Satisfaction Index. Therefore, after the patient has been treated by the recommended doctor and surveyed for this index, her overall satisfaction can be incorporated into the RSI.

The RSI is a measure compiled from the Diagnostician's recent recommendations. The table below illustrates how the RSI is computed; for simplicity we used just three of the Diagnostician's recommendations. Recommendation 1 is on The Cutting Edge and so it has a score of 1.0. The third column shows the patient satisfaction score. The average score shown in the fourth column of the table is a simple average of the Edge score and the patient satisfaction score; the RSI is computed as 100 times the average of these average scores. Thus, in this example, the Diagnostician's RSI would be equal to 100 x (.935 + .905 + .825) / 3 = 88.8.


Computation of Referral Satisfaction Index
Referral Edge Score Patient Satisfaction Score Average Score
1 1.00 .87 .935
2 .92 .89 .905
3 .80 .85 .825

The Doctor Shopper reports the Diagnostician's RSI score and the average of the RSI scores for all Diagnosticians in the local market. This allows the patient to compare the "efficiency" of his own Diagnostician's recommendations with those of the average Diagnostician in the community.

The doctors listed in The Doctor Shopper are those whose offices are conveniently located with respect to the patient's home or place of work. Thus, the chances are good that the Diagnostician would be able to refer the patient to a doctor who shows better performance than the ones listed according to this convenience criterion, since the Diagnostician can choose from among all the doctors within—or even outside of—the local market area. The Diagnostician's recommendation is listed right below the patient's diagnosis in The Doctor Shopper, and it is also listed with the other doctors to facilitate comparison. In our sample of The Doctor Shopper, Dr. Greene, the Diagnostician, has recommended that the patient receive treatment from Dr. Grant.

Diagnosticians have an incentive to achieve high RSIs, since high RSIs will make their recommendations more credible to their patients; indeed, the patient might well choose to ignore the Diagnostician's recommendation if his RSI is not favorable. As an added incentive for making outstanding recommendations, the RSI would be one criterion on which Diagnosticians are evaluated when determining their compensation.

Diagnosticians will be able to improve their recommendations by reviewing PSI scores of doctors in their local healthcare market and by consulting the relevant Value Charts. Of course, if the Diagnostician personally knows the doctors, he will be in an even better position to recommend one whose personality will mesh well with that of his patient. All Diagnosticians will want to do well by their patients, and being in a position to refer them to the best doctors for treatment is an important part of their commitment to their patients.

There is one other benefit from the RSI that we believe to be exceptional. We know that doctors welcome new referrals. If they want to increase the number they receive, they should try to position themselves on The Cutting Edge and to achieve high patient satisfaction scores. To move to the Edge, doctors will have to lower their prices and/or increase their performance rating. Needless to say, the RSI provides exactly the right incentives.

Diagnostic Accuracy Index. One of the most important functions of a Diagnostician is to establish his patient's diagnosis, and to do this in a timely manner. In fact, the Diagnostician assigns up to three possible diagnoses to the patient, as well as the probability that each is the true diagnosis. The purpose of this is to alert the treating doctor that the Diagnostician has some uncertainty about the correct diagnosis. The Diagnostic Accuracy Index (DAI) that we have developed accomplishes several objectives. First, over the long run, if the Diagnostician correctly selects the possible diagnoses and accurately assigns their respective probabilities, he will achieve a higher DAI score.

Second, the Diagnostician is also evaluated with respect to the timeliness of his Diagnosis. The more quickly he arrives at the correct diagnosis, the more quickly the correct treatment can begin, which will accelerate the patient's recovery. However, if the Diagnostician immediately orders all possibly relevant diagnostic tests for the patient, the cost will likely be unnecessarily high. Therefore, an initial period needs to be selected that is sufficient for the Diagnostician to perform an optimally sequenced test protocol. This is a "no penalty" period, and its length may also depend upon the diagnosis that is ultimately determined. Beyond this period, the longer it takes for the Diagnostician to arrive at the correct diagnosis, the more adversely will his DAI be affected. Since Diagnosticians are also evaluated with respect to the costs they incur to establish a diagnosis, the incentive for Diagnosticians is to follow the testing protocol, obtain the tests from the most cost-effective and responsive suppliers and accurately report the likely diagnoses and their most likely probabilities. This approach is most likely to result in the timely administration of the correct treatment.

Prognosis Rating Index. Another important function of the Diagnostician is to predict the likely recovery path of each patient receiving a written treatment plan. The prognosis rating (PR) assigned to each patient defines the recovery path that the Diagnostician expects the patient to follow if treated by a doctor with average capabilities. It will significantly affect the treating doctor's outcome scores, which comprise his Outcomes Index. Once a substantial database has been produced containing the prognosis ratings, predicted recovery paths, actual recovery paths and patient risk factors for individual patients, a statistical analysis will provide us with the capability of offering guidance to the Diagnosticians on the appropriate range of a prognosis rating, given an individual patient's risk factors. The Diagnostician will have the liberty to select the best prognosis rating within the given range.

Performance of Treating Doctors

Patient Satisfaction Index. The PSI is based on patient surveys, following an office visit. The questions in the survey range from how long the patient had to wait before seeing the doctor, the completeness and clarity of the information that the doctor provided to the patient, and the patient's satisfaction with the outcome itself. The PSI reported in The Doctor Shopper is the patient's overall satisfaction with the doctor, but more-detailed results will be available online. A perfect score on the PSI is 100.

Nurses Index. The motivation behind the Nurses Index is that nurses are "insiders" with good information. They have superior knowledge about the doctors for whom they work, and they also network with other nurses. It stands to reason that they will seek treatment for themselves and for their family from only the most capable doctors. When the RE·MEDI system initiates operations within a community, nurses will be surveyed to elicit the names of the local doctors they hold in the highest regard professionally. The results from these surveys will be reported in The Doctor Shopper. However, once a substantial number of nurses working in the local healthcare market is enrolled in RE·MEDI, then the insurance claims for services provided to these nurses and their family members can be flagged and tallied during processing under the QuickSafe system. After a sufficient amount of data has been accumulated, the survey results reported in The Doctor Shopper can be replaced by the actual number of different nurses and members of their immediate family who have been treated by a particular doctor in the past 12 months.

Outcomes Index. Our best performance measure—by far—is the RE·MEDI system's Outcomes Index (U.S. Patent 8,392,215). It is based on the unique structure of the RE·MEDI system. It provides an unbiased, risk-adjusted measure of how well patients recover from a specific illness/injury when treated by individual doctors. We believe that our Outcomes Index measures a doctor's performance more accurately than any other measure available in the marketplace today. The following discussion shows how the Outcomes Index is constructed.

The Diagnostician's knowledge of the patient and her illness uniquely qualifies him to issue a prognosis, which is his prediction of how quickly and completely the patient will recover if treated by a physician with average capabilities. In addition to the severity of the patient's illness, the Diagnostician will have knowledge of her age, co-morbidities and other risk factors that are likely to affect her recovery. Her outlook and personality may also affect her recovery. Evidence suggests that an optimistic person is more likely to have a better recovery than one whose outlook is pessimistic, other things being equal.

Therefore, while preparing the treatment plan, the Diagnostician assigns a prognosis rating (PR) to the patient. The PR is the Diagostician's assessment of the patient's recovery risk and is measured on a ten-point scale, with ten indicating the most optimistic prognosis. Because the Diagnostician has no incentive to assign other than the most accurate prognosis, his rating will be unbiased. Indeed, as seen above, Diagnosticians' compensation depends in part on the accuracy of their diagnoses, so they have a positive incentive to improve their ability to predict a patient's outcome. Diagnosticians should be consistent in the way they assign PRs to their patients. If any two Diagnosticians have identical knowledge of a patient and her illness, then, ideally, each would assign to her the same PR. We have developed statistical tools that will help to achieve this consistency.

From the patient's perspective, the ideal outcome is one in which her longevity is no worse than what it was prior to the onset of her illness, all negative illness- or injury-related signs and symptoms disappear immediately, and all functionality is restored to pre-illness levels. While death is an objective fact, the patient is the best evaluator of these other conditions. Therefore, periodic input from a living patient is all that is required to assess her recovery.

Once a patient receives her prognosis rating, she goes out into the marketplace and selects a doctor to treat her. Periodically, during her recovery—usually, while she is waiting to be seen by a doctor—she self-administers an electronic survey to characterize the current status of her illness- and injury-related effects (IREs). Her responses are then tabulated, and her recovery scored mechanically on a ten-point scale, with a ten indicating the quickest and most complete recovery and zero indicating immediate death. The recovery scores of patients with the same illness and similar prognosis ratings are then collected and averaged to produce an "Outcomes Index" (OI) for that doctor.

Referring to the "Performance" section of The Doctor Shopper, the Outcomes Index is reported for each of the listed doctors. These OIs, which in the current case pertain to a hernia repair diagnosis, are reported only with respect to the particular prognosis rating that the Diagnostician has assigned to the patient. This insures that comparisons are restricted to patients who have faced similar recovery risks. In the case at hand, the reported OIs are all based only on patients having a prognosis rating of around 8. Among the doctors listed, Dr. Haines is shown to have the lowest OI (68), while Dr. Thurgood is shown to have the highest (83).

It is clear that the Outcomes Index reports precisely the kind of information that healthcare consumers need to find the most effective doctors. (Learning what the index numbers mean, however, requires an additional effort by the patient. Information available from the Diagnostician's receptionist—or online—can show how the patient's expected recovery experience will change if treated by a doctor with a higher or lower Outcomes Index.)

It is possible under the RE·MEDI system to provide feedback so that Diagnosticians can "fine tune" the prognosis ratings that they assign to their patients. Once a sufficient database has been developed for local doctors, doctors who practice at an "average" level with regard to the recovery criteria can be identified. We can then identify any Diagnostician whose patients consistently recover at a better-than-expected or worse-than-expected rate when treated by these "average" doctors. When provided with this feedback, these Diagnosticians can adjust their excess "optimism" or "pessimism" accordingly.

Because healthcare consumers will undoubtedly be attracted to the better performing doctors, the Outcomes Index will cause doctors to compete with respect to performance. Indeed, we expect this competition to cause healthcare quality to increase substantially. Because the best performers can charge higher prices and earn higher incomes, doctors will be motivated to practice at the highest level (for a given resource use) and to improve their skills so that they show attractive results in The Doctor Shopper.

There is one other beneficial aspect of the RE·MEDI Outcomes Index. Patients receiving a high (optimistic) prognosis rating from their Diagnostician know that just about any competent doctor can produce an effective recovery within a reasonable amount of time. Therefore, they will tend to select a relatively inexpensive doctor, such as Dr. Williams, who will yield for them a relatively large Ca$hback Coverage payment ($830).

On the other hand, patients whose prognosis ratings are low understandably will be apprehensive about their recovery, and they surely will want to be treated by a doctor in whom they can have a high degree of confidence—a doctor such as Dr. Thurgood. To obtain an outstanding doctor in a situation as serious as their pessimistic prognosis rating implies, most patients will willingly incur out-of-pocket costs. Indeed, to be treated by Dr. Thurgood, the patient would have to pay $2,710 from her own pocket.

The RE·MEDI prognosis rating thus produces a most desirable result; namely, it tends to reserve Dr. Thurgood and other exceptional doctors for only the most difficult cases, for which their skills are most prized. Similarly, average doctors will tend to receive those cases that place modest demands on their limited skills. This is just as it should be: We do not want Dr. Thurgood wasting his time with routine cases when he could be in the operating room performing his wizardry; nor do we want the less skilled attempting to practice beyond their capacity.