The implementation costs of an electronic prevention programme in Belgian general practice

Abstract

Introduction: Guidelines to prevent cardiovascular (CV) disease are widely available. To implement these guidelines an electronic prevention programme (EPP) with a risk calculator for general practitioners (GPs) was developed. The aim of the present study was to calculate the implementation cost per installation. Methods: This cost study is part of a larger clinical trial, studying the effects of interventions in GP-practice on the management of CV risk factors. Participating GPs were asked to install the EPP. They could take part in a group education session or receive education by e-mail, telephone or at home. After a prospective cost registration, the cost per installation and a sensitivity analysis were calculated. Results: 185 GPs participated in the study. The total implementation cost of the EPP was €83,939. As the EPP was successfully installed by 102 GPs, the mean cost equals €823 per GP. Sensitivity analyses showed a decrease in costs due to a decrease of the costs of group education and/or an increase of installations.

Conclusion: This study showed that it is possible to implement an EPP for cardiovascular prevention with an acceptable cost.

Key words::

• Costs
• medica record
• revention
• general practice

Introduction

Cardiovascular (CV) diseases are the leading cause of death and the third cause of disability in Europe (1). Guidelines to prevent CV diseases are available (2). To implement these guidelines in general practice, an implementation plan has to be worked out taking into account the bottle-necks in actual care (3).

After studying the needs of the Belgian general practitioners (GPs), we developed an electronic prevention programme (EPP) with a risk calculator based on the European guidelines for CV prevention (4). The EPP calculates the risk and generates standardized therapeutic goals depending on the risk profile and stage of behaviour change. The development of the EPP is one intervention mode to support GPs and will be evaluated in a three-year clinical trial. In the latter trial the effects of multidisciplinary interventions in GP-practice on the management of CV risk factors will be studied. The GPs participating in this trial could install the EPP. After installation, the EPP is automatically linked to the existing electronic medical files (EMF) of the GP’s.

The need for information and communication technology (ICT) in health care grows and the Belgian government stimulates GPs since 2000 to use an EMF. Although 76% of the Belgian GPs use an EMF, little is known about the implementation costs of ICT tools in general practice. An economical evaluation of the implementation of ICT tools in companies shows an underestimation of these costs (5). The aim of this study was to determine the cost for installation.

Methods

Study design

This cost study is part of a registered clinical trial approved by the Hasselt University Ethics Committee (ISRCTN23940498) (4) in which 350 patients are randomised in an intervention and a control group. Both groups receive several prevention consultations by their GP using an EPP. Only the patients of the intervention group receive a follow-up with intensive support of health behaviour change via different delivery modes (i.e. a tailored website and personal advice of a multidisciplinary team) (6). The primary outcome measures will be cardiovascular risk factors. The GPs situated in the region of the included study patients were invited to participate in the study (n = 792). They received a written invitation by the GPs’ president and a telephone and e-mail from a study collaborator to participate in the randomized clinical trial. An informed consent was signed by 185 GPs (23%).

Electronic prevention program (EPP)

Figure 1 shows a screen print of the EPP. Medical risk factors and stage of behaviour change (healthy food, physical activity, not smoking) have to be filled in per patient via the input screen. The EPP calculates the cardiovascular risk and generates medical and behavioural advice with therapeutic goals (output screen) (2). The cardiovascular risk calculation is based on the patient’s cardiovascular history, glycaemia, systolic blood pressure, total cholesterol, age, gender and smoking habit using the SCORE algorithm. If a person has diabetes or has suffered from a personal ischaemic event or has a SCORE risk ≥4, he is classified as a patient with a high risk to die within 10 years from a cardiovascular event. A patient is also considered to have a high risk if he has one exuberant medical risk factor: cholesterol level ≥320 mg/dl; blood pressure ≥180/110 mm/Hg; LDL cholesterol ≥240 mg/dl. A SCORE risk 2–4 is classified as medium risk and ≥2 as low risk. Based on the input, the EPP generates an individual advice with therapeutic goals. This advice can be printed for the patient to take home.

Figure 1. Electronic Prevention Program (EPP) with risk calculator.

The risk calculator also includes a function to send each consultation and the patient’s calculated risk to other physicians using secured data transport (Figure 2). SIX (Secure Information eXchange) is a web service based platform that ensures authentication, confidentiality and message integrity through PKI (Public Key Infrastructure).

Figure 2. Data transport.

To implement the EPP, the participating GPs were invited to a group education session that consisted of a theoretical course on cardiovascular prevention and a computer training session. Afterwards, they received an e-mail with a hyperlink to download the EPP. If a GP successfully installed the EPP, the study collaborator automatically received a confirmation message. If this message was not received, the study collaborator contacted the GP and offered assistance by telephone or by e-mail. If this intervention was not successful, the study collaborator suggested technical assistance at home by an assistant. GP’s were free to accept this extra service.

Definition of costs

A societal perspective was chosen. All costs to implement the EPP were progressively registered from the 1 January to 31 December 2007. Following costs are included: programme costs (e.g. personnel, overhead, transport, education material) and costs to participants (e.g. time, transport). Overhead costs are costs for resources that are shared by different departments (e.g. power, heating). Since university rooms are for rent to others, this price, for a total of €800 is assumed to cover for the overhead cost.

For each cost type a distinction in fixed and variable costs is made. Fixed costs of a programme are those that do not vary with the level of activity (number of GPs included) (7). Variable costs are those that change as the level of activity changes (level of education, number of GPs included). The variable costs will be determined by multiplying the quantity with the unit cost (Table I).

Table I. Unit costs for material and personnel input.

Material input
Resource	Value (€/unit)
Transport	0.15 [16]
Printed colour page	0.92
Printed Page, black and white	0.03
Collate	1
Coloured paper	0.02
Stamped envelope	1
Personnel input
Resource	Value (€/h)
Psychologist	50
Consultant	150 [8]
Programmer	50
Administrative employee	27
External assistant technical support	40
General practitioner	74 [8]

The quantities to determine the personnel costs were electronically registered. Each employee involved in the study could log on to a website to register an activity towards a GP. The duration of the activities in minutes was used as a quantity for the calculation of the costs for personnel input. These quantities were multiplied with the unit costs (i.e. wages per hour) (Table I). The unit cost for GPs was calculated based on the proportion of patient visits performed at home (30%) and at the office (70%), their duration as reported in literature and the present tariffs for GPs (€31.2 for home visit and €18.1 for consult in GP-practice) (8).

The mean costs for different education methods are determined as follows: the total costs for individual education per e-mail are divided by the number of e-mails sent, the total costs for individual education per telephone are divided by the number of telephone calls, the total costs for individual education at home are divided by the number of home visits and the total costs for the group education sessions are divided by the number of those present.

The total cost of the implementation is calculated by summing all the costs incurred. The mean cost for an installation is calculated by division of the total cost by the total number of installations.

Statistics

Microsoft Excel and SPSS version 14.0 for Windows were used. To find possible differences between participants and non-participants we performed chi square tests.

To examine the influence of uncertainties in the variables, a one-way sensitivity analysis was performed varying different variables along a plausible range of values while holding all other variables constant. Owing to the prospective registration of costs, the margins for error are chosen arbitrary to fall within a range of 10%. The influences of uncertainties are examined for individual education per telephone, per e-mail, at home and for group sessions as follows costs base case –10% and 110%. The range of installations goes from half of the installations (n = 50) to a doubled number of installations (n = 200). The costs that vary with the number of participating GPs are the variable programme costs and the variable costs to participants. Both costs are divided by two (half of the base case installations), multiplied by two (doubled base case) or multiplied by 10 (1.000 successful installations). The fixed programme costs do not vary with the number of participating GPs.

Results

Study population

A total of 185 GPs enrolled in the study (148 males and 37 females). On average the GPs graduated 25 years ago (SD ± 10 years). Of this group, 99 were solo practitioners and 86 were working in a group practice. Participants and non-participants were similar with respect to graduation year (P = 0.567), solo/group practice (P = 0.119), but not with respect to gender (P = 0.001). The proportion female in the non-responders was larger than the proportion female in the study population.

Cost per installation

Individual education per e-mail included different cost types: personnel costs and participants’ time to read an e-mail. One hundred and five e-mails were sent. Consequently, the mean cost for this education method was €10. Individual education per telephone included different cost types: personnel costs, participants’ time, and telephone costs. Forty-nine telephone calls were made. The mean cost per for this education method was €13. Individual education at home included the following cost types: personnel costs, transportation costs, and participants’ time. Fourteen education sessions were given at home. The mean cost for this education method was €53. Group education consisted of the following cost types: personnel costs for preparation of the group education, personnel costs (lectures, etc.), education material, participants’ time, participants’ transportation costs and catering. 96 GPs attended these sessions. The mean cost for this education method was €247.

The total costs for the implementation of the EPP were €83,939 (Figure 3). The EPP was successfully installed by 102 GPs. The calculated cost per installation equalled €823. The cost types that contributed most to total costs were the personnel costs. Personnel costs contributed 93% to programme costs. The costs to participants mainly consisted of the time of the GPs taking part in the group education session (82%).

Figure 3. Different cost types.

Sensitivity analysis

Table II shows the results of the one-way sensitivity analysis. A 10% decrease in the costs for the group education changed the cost for installation from €823 (base case) to €800. A doubled number of installations would change the cost for installation from €823 to €610. A tenfold number of installations would change the cost for installation from €823 to €426. If the number of installations would decrease from 102 to 50, the cost for installation would increase from €823 to €1,299.

Table II. One way sensitivity analysis.

Variables	Base case cost	Cost for installation underestimation of costs overestimation of installations	Cost for installation overestimation of costs underestimation of installations
Costs for telephone education	655	824	821
Costs for mail education	1023	823	822
Costs for education at home	747	823	822
Costs for group education	23,685	846	800
Number of installations	102	1,299	610

Discussion

An electronic prevention programme (EPP) with risk calculator and linked to the electronic medical files (EMF) was installed on the GPs’ computer (2,4). In this study we calculated the different costs for installation. Hundred and two of the 185 GPs in the study successfully installed the EPP. The mean cost per installation was €823. A decrease of the costs of the group education or an increase in the number of installations can lead to the lowest possible cost per installation.

The prospective electronic registration of all costs is an advantage of this study. Another strength of the study is the fact that we based our EPP on issues found by Terry et al. These authors concluded after different focus groups that four issues have to be studied before implementing an EPP in GP-practice (9). First, ‘GPs’ expectations of the EPP and their needs to use software’ have to be studied. In our study the expectations and needs were discussed with several GPs before programming the EPP. A second issue, ‘the level of commitment of all parties to the implementation and adoption of electronic tools’ was foreseen by different levels of support by the study team. The third issue, ‘the leadership of the construction and of the implementation of the EPP’, was performed by the study coordinator. Finally, the fourth issue, ‘the knowledge of computers potential of the EPP users’ was anticipated by an education session with practical workshop for the installation and the use of the EPP.

This study, however, has several limitations. First, the individual education at home was only given on GPs’ demand. A more proactive offer to install the EPP at home could have been more successful. Furthermore, this self-selecting mechanism resulted probably in the fact that only highly motivated GPs demanded individual education at home. A second limitation is the absence of data on the use of the programme in daily practice of the GP. A third limitation is the absence of a systematic registration of the reason why 83 out of 185 GPs did not successfully install the ICT-tool. Systematic registration of obstructions could give information on the implementation or the ICT-tool itself and has to be considered in a following study. A fourth limitation is the fact that these GPs participated in a larger clinical trial, studying the effect of multidisciplinary interventions on the follow-up of CV risk factors within their patients. This could give an inclusion bias, for example, GPs were motivated to accept new tools to support their prevention management in daily practice.

The total cost of the implementation of the EPP was €83,939 with a mean cost per installation of €823. Our cost for group education, i.e. €247, was low in comparison with literature, where amounts ranging from €505–€702 were reported (8,10). The difference in costs in these studies could be explained by the higher number of education sessions. Group education is more expensive than individual education at home, mean cost of €53. This difference could be explained by the limited time during education at home (10–15 min) instead of (2 h 30 min) during the group education. The group education session, however, gives the opportunity to convince GPs of the importance of using an EPP.

It has been reported that the most cost-effective implementation method for alcohol screening in general practice was practice-based training and a support telephone call every two weeks (in comparison to practice-based training only) (11). In our study on cardiovascular prevention we did not find any benefit of assistance by telephone. In comparison to the cited study, we have not called systematically every two weeks. A more intensive support could have been beneficial.

The sensitivity analysis demonstrated that a doubled number of installations would decrease the cost per installation to €610. If the programme could be implemented in 1000 GP practices the cost per installation would drop to €426. This decrease of costs is due to a decrease of fixed costs like the development of the EPD.

The costs for GPs’ transport and contact time amounts to €20,171 for 185 GPs. However, the GPs in our study were not paid. When implementing the EPD on a larger scale the question will arise ‘who will pay these costs?’ One possibility is that the public health care organization pays. The literature reported a return on investment of $86,400 (€55,616) on five years per GP from using an electronic medical file to the health care organization due to an improved utilization of test (12). Health information technology increases the efficiency of health care (13). Another possibility is that GPs invest. Their return on investment is estimated at $14,055 (€9,047) annually, mainly due to a reduction in chart pulls (63%) (14).

Conclusion

The mean cost per installation for an electronic prevention programme on cardiovascular disease was €823. Implementation costs decreased with a higher number of successful installations. Adequate funding has to be foreseen to implement quality improving electronic tools on a large scale in general practice.

Acknowledgements

This study was funded by the ‘Leerstoel De Onderlinge Ziekenkas-Preventie’ The sponsor had no involvement in the study design, the data collection and analysis, the manuscript or the submission. The authors should like to thank Pfizer Belgium for offering financial support for the technical development of the EPP. Pfizer Belgium was not involved in the study design, the content development of the EPP or in writing and submitting the manuscript.

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.