IMPROVING THE REFERRAL PROCESS

Changing the System to Boost Satisfaction

Referrals ranked among the top three reasons for complaints at this health plan in 1996. Faced with an industrywide trend to relax referral requirements, and a high complaint rate regarding the referral system, the plan commissioned a task force to look at reasons for low satisfaction.

Many new members in this growing plan previously had been covered by indemnity insurance. Unfamiliar with managed care rules, members seeking specialty care expressed dissatisfaction.

While the referral system is an important satisfaction issue for all members seeking specialty care, it can be a basic access issue for patients with chronic disease. Patients requiring multiple repeat specialty visits are most likely to experience problems with the system; these problems may mean delayed or deferred care.

Members were not the only ones voicing their displeasure with the referral system. The results of the plan's utilization management satisfaction survey indicated that there was widespread provider dissatisfaction as well.

With the potential financial ramifications of inappropriate or excessive specialty care use, the plan needed to proceed carefully to improve satisfaction, while still managing care.

The plan targeted all commercial and Medicare members for this activity. Any member for whom a referral was requested was part of the target population. The plan used measurements from three sources.

The plan used its HEDIS member satisfaction survey of commercial members. It looked at the results from two questions:

• "Have delays in your medical care while you waited for approval by your health plan been a problem?"
• "Has not being able to get a referral to a specialist that you wanted to see been a problem?"

The measures used were the percentage of members answering "not a problem" to these two questions.

It used complaint information from the member services database as a second source of data. Every complaint was given a reason code, including a specific one for referral problems. The plan tracked the referral complaint rate per 1,000 commercial and Medicare members on a monthly basis.

Finally, a sample of 150 referrals per month was examined for turnaround time. By querying the prior authorization system's date fields for receipt and authorization decisions, the plan tracked the percentage of requests processed within 14 days. It calculated this number quarterly.

TRACKING PERFORMANCE INDICATORS Quantifiable measurements in a quality improvement activity frequently cover an annual time frame. In this way, rigorous (and oftentimes consuming and costly) measurement gives a broad view of the overall impact of interventions throughout the quality improvement cycle. But more frequent reporting of measurements can assist in managing ongoing operations. With regular reporting of complaint and referral turnaround time data, the QI committee had the information to draw conclusions about performance. They could then track performance on a periodic basis, and take specific action if performance fell below goal.

In 1997 the plan added another measure, using disenrollment survey data from Medicare members. It looked at the percentage of disenrolling members stating that "problem with a referral" was a reason for disenrollment.

The plan used 1996 data for its baseline measurement. It selected 85 percent goals for the two survey measures, based on industrywide data on satisfaction. It decided on a 20 percent reduction in the complaint rate as a goal. Knowledge of industry performance led to the adoption of a goal that 90 percent of referrals be processed within two weeks.

A task force dedicated to service improvement reviewed employer group surveys, looking at narrative responses concerning the referral system. A committee made up of network office staff personnel provided input in identifying two major barriers to satisfaction:

• The requirement for preauthorization of all specialist consults, follow-up visits and in-office procedures resulted in delays to members
• A requirement that specialists refer members back to their PCPs for routine diagnostic services was cumbersome

At the same time, the plan convened a team to look at operational processes related to delays in referral turnaround times. This team identified a number of problems:

• The company's own internal mail system delayed the outgoing hard copy of the authorization form to members
• Providers batched requests and only faxed them in once per week
• The utilization management (UM) department fax machines were unable to handle the volume

The plan redesigned its referral process with two major changes:

• PCPs could now provide their patients paper referrals without specific plan prior authorization
• Specialists could now provide referrals for routine diagnostic services

It took over four months to implement the redesigned process. New forms were distributed to all PCPs and specialists. The plan held educational meetings at various sites to educate practitioners and their office staff on the changes. Information packets were sent to the offices following these meetings. Plan staff visited practitioner offices to follow up and answer any questions.

The first remeasurement was based on 1997 data. Disenrollment survey data from Medicare members was used for the first time.

Although neither of the two survey questions reached the 85 percent goal, both showed small increases. The proportion of commercial members reporting no problem in getting referrals for a desired specialist showed a statistically significant increase.

The 57.5 percent decrease in complaints exceeded the plan's goal for a 10 percent reduction, and was statistically significant.

The percentage of referrals meeting the turnaround time standards exceeded the plan's goals for the second and fourth quarters. Overall, there was no significant difference in performance on this measure compared to baseline.

BEST PRACTICE SHARING The plan conducted a survey of providers with high satisfaction scores on practitioner profiles. It sought to uncover what contributed to good satisfaction with the referral process. These "best practices" data were used to identify causes of satisfaction. Together with the barrier analysis, it helped the plan accomplish a major revision of the referral process.

Further understanding of the revised referral process led to the identification of additional barriers to satisfaction:

• Members wanted direct access
• Members complained that it was time-consuming and inconvenient to pick up a referral from the PCP's office
• There was perceived lack of access to specialists at a local university
• Members had a negative perception of the managed care process

The plan revised its contract language to allow PCPs to mail referrals to members or fax them directly to specialists.

It revised its relationship with the local university to allow PCPs to refer to these specialists without preauthorization.

It revised the member handbook to better assist members in understanding the referral process.

The second remeasurement was conducted using 1998 data:

Neither of the two survey questions reached the goal of 85 percent satisfaction. The small increase over baseline in the number of commercial members not reporting a problem with the waiting time for a referral was statistically significant. The decrease in percentage of commercial members reporting no problem in getting a referral to a desired specialist was not statistically significant compared with baseline.

The rate of complaints due to referrals continued its downward trend. The 43.8 percent decrease between 1997 and 1998 exceeded the plan's goal for an annual reduction, and was statistically significant.

No statistically significant trend was seen in meeting UM turnaround times, but the plan met its performance goals for the second, third and fourth quarters.

The Medicare disenrollment survey showed a statistically significant decrease 0f 13.7 percentage points.

The plan has continued to monitor the activity's performance indicators. The redesigned referral process remains in place. The plan eliminated the need for preauthorization for services such as dialysis, chemotherapy and nuclear medicine testing. Repeat measurements in 1999 showed sustained improvement. Seventy-seven percent and 76.7 percent of commercial members reported no problems respectively, with authorization delays or in getting desired referrals. Referral complaints have declined further to 0.16/1,000 members.

The plan continues to modify the referral process, increasing direct access. People with diabetes now have direct access to ophthalmologists, and women can access obstetricians directly for prenatal care.

The plan continues to modify the authorization system for particular clinical situations as part of its commitment to member satisfaction.