Cost Effectiveness Analysis of Behaviour Change Intervention: A Proposed New Approach and an Application to Egypt

Robinson and Lewis introduce a new methodology for performing cost analysis evaluations of family planning programmes. The measured output of the new methodology is the behavior change of programme clients rather than units of contraceptive services delivered. The authors introduce the new framework, test it against a case study of past communication campaigns in Egypt, and derive the cost per unit of behavior change.

Background

Cost-effectiveness analysis is an evaluation tool used to estimate the cost per unit of achieving programme outcomes and maximize performance at least cost. A typical measure for these cost-effectiveness evaluations is the couple-year-of-protection (CYP). CYP assigns each type of contraceptive a number of years of protection rating (e.g. an IUD = 2.6yrs protection), and measures the amounts of each distributed to produce a total CYP figure. Evaluation of campaign effectiveness is based on the increases in the CYP. It is a supply measure rather than a measure of actual behavior as is the alternative methodology being proposed by the authors.

The basis for their new model is the understanding within the Behavior Change Communications (BCC) paradigm that communication activities disseminate contraceptive knowledge and the attitude changes needed to influence behavior. Determining the optimal expenditures on BCC programming by calculating the cost per client-impact amount can be done if the fiscal cost and actual impact of client behavior can be determined. Being able to assess this cost more accurately would allow communication managers to make optimal usage of their limited budgets, and determine which their least-cost and most effective campaigns are.

The authors provide a substantive overview of the communication programming that took place in Egypt during the period 1978-1998, and focus specifically on the period 1995-1998, the period from which they draw their 4 case studies. Data is presented on the impacts of the overall BBC programme effort in Egypt during this period. Substantial gains have been made in Egypt, including a decline in the 'ideal family size' and an increase in contraceptive demand. The four specific family planning BCC mass media campaigns selected as case studies are:

• The 'Ask-Consult Campaign' (1995-1997): a private sector initiative aimed at improving interaction between private sector pharmacists and their prospective clients.
• The 'Gold Star' campaign (1996-1998): A logo system that was introduced with private sector cooperation as a certified quality-of-care indicator to improve clinic standards and encourage public usage of authorized providers.
• Dr. Karima Mukhtar campaign (1995-1996): A series of dramas in which a female doctor answered various questions about family planning and provided information on the newly introduced injectable contraceptives.
• Ammed Maher male responsibility campaign (1995-1996): that was aimed at encouraging men to take a greater role in family planning and promoted condom usage.

Methodology

The costs of each of these BCC programmes are known and detailed books have been kept. The production, broadcast and total costs in Egyptian £s (E£) are presented for each campaign in a detailed table. Estimating the campaign specific impacts of the BCC projects is much more challenging. The authors begin by identifying one main particular form of behavior change from each campaign - the use of a particular type of contraceptive. Each of the four campaigns emphasized a different contraception option.

The relevant indicators of impact are (1) the number of airings of each campaign per year, (2) the number of persons reached, and (3) the average number of exposures per person - which allows for the determination of total exposures for each campaign. The authors then calculate the cost and reach figures into cost per client reached and the cost per client-exposure for each of the four campaigns. The authors are not specifically interested in the effectiveness of each campaign, but merely the reach.

The following is a representation of the formula presented by the authors.

(Production Costs) + (Airtime Costs) = (Total Costs)

-->>

(Number of Persons Reached) X (Average Exposure) = (Total Exposures)

-->>

(Total Costs)/(Total Exposures) = (Cost per single-client exposure)

-->>

Results

This evaluation data from each respective media campaign and the results of Robinson and Lewis' calculations are as follows (*at the time of the study 20 £E = 6 $US):

The 'Ask-Consult' campaign:

• Recalled by 70% of respondents who knew that the pill was available at pharmacies, an estimated 3.5 million women were supplied information about the pill.
• An estimated 83% of the target population reached. However, pill demand from 1995-97, remained steady at 70%.
• Cost Analysis: Cost per airing = 8688 £E s; Cost per person reached = 0.29 £Es, Cost per single-client exposure = 0.02 £Es.

The 'Gold Star' campaign:

• 81% of the target population was reached for an estimated total 8.1 million.
• Associated with 900,000 net new IUD users, an increase from 30-35% of the population over 1995-1997.
• Cost Analysis: Cost per airing = 4730 £E s; Cost per person reached = 0.18 £Es, Cost per single-client exposure = 0.01 £Es.

Dr. Karima Mukhtar campaign:

• 70% of the target population was reached for an estimated total of 7 million.
• Associated with an increased use of injectables from 2.4% to 3.9%, a net increase of 230,000
• Cost Analysis: Cost per airing = 5440 £E s; Cost per person reached = 0.09 £Es, Cost per single-client exposure = 0.01 £Es.

Ammed Maher male responsibility campaign:

• 45% of the target population reached for an estimated total of 4.5 million.
• Condom use increased only from 1.4% to 1.5% during the study period.
• Cost Analysis: Cost per airing = 7358 £E s; Cost per person reached = 0.49 £Es, Cost per single-client exposure = 0.03 £Es.

The authors' final cost calculations for the set of campaigns suggest that it cost about 10£E (approx. 3 $US) to give a client the knowledge and motivation needed to lead them to adopt contraceptives. The authors believe that this figure is complementary to the CYP model for cost estimation which estimated that it would require 20£E and argue that this adds credibility to their proposed methodology.

Conclusions

The authors do not seek to suggest that one media campaign was substantively more cost-effective than another, but do note that the four separate campaigns each had an effect on contraceptive usage. More importantly, however, is their belief that this study helped to demonstrate the workability and utility of their model for calculating cost per client-impact. They suggest that this requires that a specific behavior change, rather than general exposure must be identified for each evaluation, and that a considerable amounts of data on coverage rates, demand for products or services, and financial costs must be available for such analysis. They argue that the type of data needed to perform these types of cost-per-unit assessments could easily be added to DHS and impact surveys.