CERVICAL CANCER SCREENING

Improving High-Volume Screening in an Expanding Population

Mortality from cervical cancer has declined by 80 percent since the introduction of the Pap smear in the early 1950s. This has been one of the all-time successes in cancer control. But the rate of decline in cervical cancer mortality has slowed since the 1970s. In 1993, 13,000 American women developed cervical cancer, and 4,400 died. [1] Most of these deaths could have been prevented if we could simply screen all women. [2]

Over one-third of the total membership of this HMO was made up of women ages 21 to 64. With a large and growing female population, and the documented effectiveness of the Pap test in preventing death from cervical cancer, the plan was chagrined to find its early HEDIS measurements to be substantially below benchmark. It decided to undertake a formal quality improvement activity to increase its cervical cancer screening rate.

The plan used an audited HEDIS 3.0 measure for cervical cancer screening for this activity. It used administrative (claims/encounter) data. No sampling was used; the plan targeted the entire population of women age 21 to 64 who met HEDIS criteria. The measure looks at the percentage of continuously enrolled women who had received one or more Pap smears in the last three years.

Baseline measurement, from 1997 data, showed a rate of cervical cancer screening of 43.4 percent. The plan identified a benchmark of 83 percent, from the 90th percentile in NCQA's 1997 Quality Compass. It set a performance goal of 78 percent by the year 2000.

A multidisciplinary work group that included four QI department personnel, the clinical operations director, the director of utilization and case management, a health educator, and the medical director supported the activity. The plan's corporate parent supported the activity by supplying study design and statistical analysis through a subsidiary dedicated to these functions.

SCARCE RESOURCES Even though the plan could allocate no new resources to the project, it felt the imperative to proceed with this activity. The work group identified ways to use existing resources. Additional tasks were incorporated into the workloads of existing personnel. To stretch resources, it combined member education and reminder mailings for cervical cancer screening and breast cancer screening. It used internally developed educational materials, rather than purchasing "off-the-shelf" printed brochures.

The work group performed a formal root cause analysis. Using a fishbone diagram (included in QP Tool ), it identified barriers of perceived importance. Interventions were designed to address these barriers. Physician members of the QI Committee helped analyze the barriers and plan interventions. The work group then reviewed the results of a 1995 survey of noncompliant members (included in QP Tool ) to assess members' perceptions of these barriers. Interventions to address barriers were prioritized based on the frequency of mention of each issue by survey respondents.

The top barriers to be addressed were:

• Lack of member awareness and misinformation
• Perceived discomfort of Pap smear
• Inconvenience of the scheduling
• Forgetting to get the exam
• The member's physician did not recommend the exam

The plan identified major opportunities to overcome the top barriers:

• Member education to address awareness and discomfort
• Identifying noncompliant patients to their physicians
• Informing practitioners of the importance of advising their patients to have the screening
• Directly reminding noncompliant members to have the test

The plan sent targeted mailings to noncompliant members. These consisted of:

• A signed letter from the medical director
• A fact sheet about cervical cancer and Pap smears
• A card outlining a gift incentive for getting a Pap smear

(The mailings are included in Appendices 28, 29, and 30.)

MEMBER INCENTIVES The plan implemented an incentive program through which women could get a $20 gift certificate good at one of four stores if they submitted a postcard signed by their physician documenting the exam. Although novel, the plan ended up abandoning this approach. It made the not uncommon discovery that such programs are difficult to administer. Although the plan faxed physician offices with information about the program, it found that getting participating physicians to sign and complete such cards could be problematic.

The plan also distributed pamphlets stressing preventive health to its members and published member newsletter articles. It sponsored community-based free and low-cost screenings, as well as a PBS special on women's health that was shown locally.

The plan sent all primary care and OB/GYN physicians preventive care recommendations, guidelines and information about Pap tests.

A targeted mailing, to physicians of noncompliant women, included the patients' names and information about the importance that members place on direct recommendations from their physicians. The mailing included information about the plan's expanded coverage of Thin Prep Pap smears.

The first remeasurement, based on data from 1998, documented a rate of 65.1 percent, an increase of over 20 percentage points, and a statistically significant improvement (chi square = 862.2, p<.0001).

The plan had not yet met its year 2000 goal, and continued to work on improving cervical cancer screening.

The plan has continued to uncover additional barriers. It noted that 54 percent of the noncompliant women were age 35 to 51. An OB/GYN physician on the QI Committee noted that women of this age are less likely to have regular visits for either birth control or hormone replacement. The plan decided to focus future interventions on this low compliance group. It found a similar need among women with partial hysterectomies.

It has also focused on interventions to improve access, including extended hours and expanding the physician network, especially with regards to female physicians.

It modified its provider mailings to provide biannual physician-specific profiles with comparative peer performance in addition to lists of noncompliant members.

Outreach efforts to patients have continued. The plan found that the amount of time to manage the project decreased substantially, because many of the materials and processes had already been developed.

PERSONALIZED REMINDERS The plan learned several key points about sending reminders to its members. It noted that after deciding on an intervention, it can take three to four months to send out a mailing. Mailings themselves need to be non accusatory in tone. Personalizing the reminders helps, especially including a staff person's name and phone number. Members do call for more information, to ask questions, or to give valuable feedback. Finally, repeating the same message multiple times can have substantial impact.

The 1999 cervical cancer screening rate was approximately the same at 63.2 percent. The plan has shown it could sustain its previous meaningful improvement, but must continue to overcome more barriers to meet its year 2000 goal, as well as the ultimate objective: to see that no women die from this highly preventable disease.

[1] - American Cancer Society, Cancer Facts and Figures-1993 (Atlanta, GA: American Cancer Society Incorporated, 1993), 13.

[2] - Department of Health and Human Services, Centers for Dis ease Control and Prevention, The National Strategic Plan for Early Detection and Control of Cervical Cancers. (Atlanta, GA: Department of Health and Human Services, Centers for Disease Control and Prevention,), 7.