Cost-effectiveness of a mailed educational reminder to increase colorectal cancer screening

To test the effect of a mailed educational reminder on increasing FOBT card return rate for CRC screening, we conducted a double-blind, randomized controlled trial in a U.S. Veteran patient population. Patients meeting inclusion criteria (age ≥ 50, not up to date with CRC screening) were randomly assigned to receive either usual care or usual care with a mailed educational reminder (intervention), which was mailed 10 days after the patients received their FOBT cards from the laboratory. All patients were given 6 months to return the FOBT cards.

We evaluated patients from three VA primary care clinics in San Diego and Vista, California between June 1 and September 9, 2007. The study included asymptomatic men and women age 50 years or older who agreed with screening and received FOBT card kits with a postage paid return envelope. Under usual care, primary care physicians (PCP) entered a computerized order for FOBT for CRC screening, and then patients were instructed by their PCP to pick up FOBT cards from the laboratory and return them for analysis. Patients were excluded from this study if they were less than 50 years of age, were currently on an inpatient unit, were up-to-date with CRC screening, or refused to undergo any routine CRC screening. We did not exclude any patients over the age of 75 as recommended by the USPSTF in 2008, because our study was performed prior to the 2008 guidelines [6].

The mailed educational reminder consisted of an 8.5 × 11 paper folded in thirds, personalized, sealed, and sent to the subjects' home address 10 days after the patients were given their FOBT cards by the clinical laboratory (Additional file 1, Appendix 1). The mailed letter was one-sided, written at an eighth-grade reading level [21], and contained a reminder to return their FOBT cards on the top portion of the letter. The middle portion of the letter had several statements regarding the risk of developing CRC, who is affected by the disease, and the benefits of getting screened. On the bottom third of the letter was a quote from a United States Veteran colon cancer survivor, who emphasized the importance of colorectal cancer screening. The Institutional Review Board (IRB) at our institution granted exemption of an informed consent because of the minimal risk associated with the mailed reminder and a guarantee of at least usual care for all patients.

An analysis of costs was conducted from the payer's perspective to help better represent the costs that a healthcare system may incur when offering interventions of this type. The payer's perspective will allow healthcare organizations to gauge approximately what it would cost for them to conduct a similar program in the future; assuming adjustments are made for inflation. Intervention costs were based on the actual personnel time and materials used in the mailed reminder and are detailed in the next section. All research related activities were excluded from the cost analysis.

To calculate the costs of our intervention, we identified all inputs used for the intervention. The cost of our intervention included the cost of creating the letter content by a physician, editing and approving the content by management (Gastroenterology (GI) section chief), the time to generate and personalize the mailed reminders by an administrative assistant, overhead costs, and finally the cost of the mailed reminder itself including paper, envelopes, pen, printer ink, and postage (Table 1). Personnel cost estimates were derived from actual salary and benefit data, which was $40,000 per year for an administrative assistant, $150,000 per year for a physician, and $200,000 per year for a VA staff gastroenterologist. The labor costs were calculated by multiplying the time spent of performing the task (creating the letter content, editing and approving the content, generating, personalizing, and mailing the reminder) by the employee's wage per hour. We determined the costs of all materials used for the mailed reminder by multiplying the costs of each material by the number of patients receiving the mailed reminder. Overhead costs were estimated at 69% of the personnel costs required to deliver the intervention, which accounts for facilities costs, indirect support personnel, and other typical indirect costs associated with running an outpatient healthcare program [22]. This figure of 69% is based on data showing that only half of all healthcare reimbursement costs are related to direct provision of care and that roughly 69% of non-care costs are indirect costs [23, 24]. This method has been used in the cost analyses of other similar trials [25]. Other costs, such as those associated with research and development (R&D) activity were excluded. First-copy costs, defined as costs incurred in establishing an intervention, are considered quasi-fixed costs independent of the number of units produced once production is started [26]. In general, first-copy costs are excluded when they involve situations where much of the intervention is already in existence and only modification is needed to adapt it for implementation [14]. This is assuming that the final product can be made available to other health care organizations through public access [14]. Furthermore, it is expected that managed care organizations or community clinics would not develop their own mailed reminder but would rely on public access or licensing arrangements.

The incremental cost-effectiveness ratio (ICER) was derived based on the difference in the costs assigned to the intervention and control groups divided by the difference in FOBT card return rates between the intervention and control groups. The result is a measure of efficiency, reported as dollars per additional patient screened. The lower the result (ICER), the more efficient is the intervention. Sensitivity analysis estimates were based on a 10% and 25% variation around the point estimates for the various cost inputs (i.e. labor costs, material costs, the effectiveness of the intervention) [27]. In addition, the ICER was similarly calculated for various subgroups including gender, race, age, current illicit drug use, current alcohol use, current tobacco use, psychiatric disorders, and history of prior FOBT completion for CRC screening. Lastly, we reviewed the literature and compiled a list of other ICER for patient-directed CRC screening promotion interventions.

Among 846 patients who received FOBT cards for CRC screening during our 3-month enrollment period, 769 were determined eligible and were randomized to receive either a mailed educational reminder (n = 387) or usual care (n = 382). The baseline demographic and clinical characteristics were well matched in both the control and intervention groups [12]. The mean age of our study participants was 63.1 years (SD = 9.6). The majority of our study participants were male (96.3%), Caucasian (72.8%), and not married (56.6%); the largest minority group in the cohort was African-Americans (11.7%). Many members of the cohort were currently drinking alcohol (45.0%) and smoking tobacco (27.6%), but only a few were noted to be currently or recently using illicit drugs (6.5%). Major psychiatric disorders present in our cohort were anxiety disorders (11.3%) and mood disorders (28.9%).

The proportion of patients who returned the FOBT cards was significantly higher in the mailed reminder group than those who did not receive a reminder. At 6 months after card distribution, the FOBT card return rate in the intervention and control arms were 64.6% (250/387) and 48.4% (185/382), respectively (P < 0.001) (Figure 1). The percent incremental effect was 16.2 percentage points and the absolute difference in FOBT card returned was 65. The timing of the return of FOBT kits was previously published, and the data indicated that 39-49% of patients returned kits by 30 days in the usual care and intervention control arms, respectively; and this increased to 47-62% by 90 days and to 48-65% by 180 days [13].

The total cost of the intervention was $962, which is $2.49 per patient (Table 1). Our startup costs, which consisted of labor costs of creating the letter content by a physician and editing and approving the content by management (GI physician) came out to be $42. Our maintenance costs were $920, and consisted of labor costs from the administrative assistant who generated and personalized the mailed reminders, overhead costs (69% of personnel costs), and the cost of the mailed reminder itself including paper, envelopes, pen, printer ink, and postage. The time spent by a physician on researching and creating the content of the letter was about fifteen minutes. The time spent on editing and approving the content of the reminder was about fifteen minutes. Lastly, the time spent on generating, personalizing, and mailing the reminder was estimated to be around 3 minutes. Sensitivity estimates based on a 10% variation around the cost inputs revealed a total cost of the intervention ranging from $865.80 - $1058.20. Note that patient-related calls to administrative or nursing staff were not measured, but may have contributed to maintenance costs and the 10-25% variation estimates.

The incremental cost-effectiveness ratio (ICER) for the mailed educational reminder (intervention) was $15 per additional person screened (Table 2). The ICER was obtained from the difference in costs per patient between the intervention and control groups ($2.49) divided by the difference in FOBT card return rates between the intervention and control groups (16.2%). Since costs included in this intervention may vary in different locations, we performed a sensitivity analysis for costs based on a 10% to 25% variation around the cost inputs (Table 1 and Table 3). Sensitivity estimates based on a 10% variation revealed that the estimated cost per additional patient screened for CRC ranged from $13.50 to $16.50. Sensitivity estimates based on a 25% variation revealed that the estimated cost per additional patient screened for CRC ranged from $11.25 to $18.75. For comparison, ICERs for other published patient directed interventions for CRC screening are listed in Table 4. In order to determine how differences in screening compliance in different patient subgroups affected the cost-effectiveness, we performed a subgroup analysis. As depicted in Table 3, the intervention had a similar ICER, ranging from $5.65 to $27.07 per additional person screened across various subgroups of patients. Interestingly, minorities such as African-Americans and Asians appear to have the lowest ICER in our subgroup analyses for reasons that remain unclear. Further studies are needed to help understand and confirm this finding.