Family Practice Vol. 16, No. 4, 353-359
© Oxford University Press 1999
Anxiety amongst women with mild dyskaryosis: costs of an educational intervention
Department of Social Medicine, University of Bristol,
a Division of Primary Care, University of Bristol and
b Division of General Practice, University of Wales College of Medicine, UK.
Kate Baxter, Department of Social Medicine, University of Bristol, Canynge Hall, Whiteladies Road, Bristol BS8 2PR, UK.
Baxter K, Peters TJ, Somerset M and Wilkinson C. Anxiety amongst women with mild dyskaryosis: costs of an educational intervention. Family Practice 1999; 16: 353–359.
Received 16 November 1998; Revised 17 March 1999; Accepted 29 March 1999.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodResultsDiscussionSummary: key points for...References |
|---|
Background. A randomized controlled trial in primary care investigated whether a structured educational intervention had an impact on the psychological morbidity associated with a 6-month period of surveillance for mild dyskaryosis. In the context of high levels of sustained distress, and few differences in terms of objective measures of anxiety, the intervention led to a greater proportion of women who were comfortable with a 6-month interval before their next smear test.
Objective. The aim of this paper is to evaluate the implications to general practices and the NHS, in terms of both costs and numbers of patient contacts, of a change from current policy to one of actively inviting all women with mild dyskaryosis to consult the practice nurse for the intervention.
Methods. We conducted a pragmatic, cluster-randomized controlled trial, comparing the intervention with standard care. The setting was general practices in Avon and South Glamorgan, UK. The subjects were women under surveillance following their first ever mildly dyskaryotic cervical smear result. The main outcome measures were as follows. Costs were reported according to randomization group, from the viewpoint of general practices and the NHS. The main elements which were costed were those attributable to production of the package and training in its use, and the costs of consultations subsequent to the woman receiving her smear test result. In addition, since in practice the intervention might be applied in different circumstances to those prevailing in the trial, a sensitivity analysis was performed to assess the costs of the educational package as realistically as possible.
Results. Almost twice as many women in the intervention group compared with the control group visited their practice to discuss their result. From the perspective of the practices, a change from current policy to the intervention policy led to potential (negligible) savings of around £3.50 per partner per year. From the NHS perspective, the intervention would lead to slightly increased costs of between £1000 and £2500 per year for an area performing 60 000 tests per year.
Conclusions. It is both feasible and acceptable for practice nurses to deliver the educational package. Moreover, from the perspective of a practice, the policy is effectively cost-neutral. The main implication for general practices is the change in the pattern of care provided: fewer women consulted their GP about their smear result and many more, following active encouragement, consulted the practice nurse.
Keywords. Economic evaluation, educational intervention, mild dyskaryosis, primary care..
|
Introduction |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionSummary: key points for...References |
|---|
Since 1992, the United Kingdom National Health Service (NHS) Cervical Screening Programme's policy has been to place women with mildly dyskaryotic cervical smear results under surveillance for 6 months before carrying out a repeat smear test. If dyskaryosis persists, women are then offered colposcopy. Surveillance of women with smear results of this nature is not unique to the UK.1 Since mild dyskaryosis occurs at a rate of about 1–3 per 100 smear tests, a substantial number of women are placed under surveillance in the UK. This policy has potentially considerable implications for primary care, as the GP and the practice nurse assume responsibility for caring for women with mild dyskaryosis.
There is evidence that receipt of a dyskaryotic smear result is initially distressing,2,3 and there is some evidence to suggest that psychological distress persists throughout the surveillance period.4 A randomized controlled trial was therefore conducted to investigate whether a structured educational intervention could have an impact on the psychological distress associated with surveillance for mild dyskaryosis.5 The trial was conducted in two health authority areas, with the approval of the four Ethics Committees. The intervention comprised an educational and risk-communication package delivered by practice nurses in face-to-face consultations, designed to inform women and to reduce their anxiety. The trial compared levels of distress and costs between a group of women receiving the intervention and a control group receiving standard care (the latter varying both between and within practices). The proportion of women who preferred the repeat smear test to be sooner than 6 months was statistically significantly higher amongst controls (74 versus 53%; 95% CI for difference 9–33%). State-anxiety scores were also markedly raised amongst control women who visited their practices and intervention women who did not (relevant interaction P = 0.009).5 Otherwise, the results of the effectiveness of the intervention show few differences.
Randomized controlled trials should measure costs as well as effectiveness, and preferably report marginal costs in order to inform policy. This paper describes the implications to general practices and the NHS, in terms of both costs and numbers of patient contacts, of a change from current policy to one of actively inviting all women with mild dyskaryosis to consult the practice nurse. In addition, since in practice the intervention might be applied in different circumstances to those prevailing in the trial, a sensitivity analysis was performed to assess the costs of the educational package as realistically as possible.
|
Method |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionSummary: key points for...References |
|---|
Although the costs which result from women being under surveillance after a mildly abnormal smear test result fall on women as well as the NHS, this paper reports costs from the viewpoint of general practices and the NHS only. Costs are reported according to randomization group.
Package and training costs
The educational package (which consisted of an A4 folder with 12 pages of information for the practice nurse to use and a condensed A5 version for the woman to take home) was developed and produced specifically for the trial;4 therefore, the cost was incurred as part of the trial running costs and did not fall on the NHS. However, for completeness in comparison of the proposed new policy with current policy, production costs are included in the analysis as a one-off payment for practices or health authorities in the intervention group. The actual cost of printing for the trial is used in the main analysis. The lowest quotes for high and low volume production, obtained from approaching a number of commercial printers, are used in the sensitivity analysis.
Prior to her first study consultation, each intervention group practice nurse was visited by a member of the trial team who provided her with a copy of the educational package and training in its use. A post-consultation questionnaire was returned by intervention nurses after consultations with study patients. This questionnaire included questions asking how long the original training session had lasted and what activities would otherwise have been undertaken in this time. The costs of nurse training sessions were calculated using these data.
Consultation costs
The small number of extra consultations per practice per year from a change to the proposed new policy is expected to have no impact on practice costs such as buildings and administrative staff. Therefore, consultations were costed in terms of nurse or GP time only.
The lengths of women's consultations with intervention nurses were obtained from the post-consultation questionnaires. The study design meant that it was not possible to collect corresponding data from nurses in the control group. However, information on the time women spent at the practice when attending either a nurse or GP to discuss their smear result was collected by questionnaire from all trial women. Nurse/patient contact time in the control group was therefore estimated by assuming the same waiting time in both the intervention and control groups, then subtracting this from the overall time spent at the practice. GP/patient contact time for women in both trial arms was again estimated from women's questionnaire data. Costs per minute of GP and nurse consultation time were taken from Netten and Dennett.6
Sensitivity analysis
A simple (single- and multi-scenario) sensitivity analysis attempts to increase the generalizability of the findings7 by systematically altering the following parameters (individually and all three combined):
- (i) the number of consultations to GPs made by women in the intervention group;
(ii) the pre-consultation preparation time of practice nurses in the intervention group; and
(iii) the costs of the package and associated practice nurse training.
Incomplete data sets dictated that some costs were calculated using averages across women. As a result, statistical tests and confidence intervals were not carried out. Any uncertainty in the mean costs in terms of statistical imprecision is considered secondary to the practical variations which are accounted for in the sensitivity analysis. In each scenario, therefore, the trial data are altered in a logical manner to reflect implementation of the intervention policy in practice.8
Scenario 1. . The arrangements in the trial were different to those anticipated in clinical practice. In the trial (where consent had first to be obtained) the invitation to consult with the practice nurse was sent some weeks after receipt of the smear result. In practice, the two could be given simultaneously—in other words, the invitation to see the nurse would pre-empt any consultation with the GP, potentially reducing consultations with GPs to zero. Of course, this may be accompanied by a rise in the number of consultations with the nurse. Scenario 1 assumes that all 73% of women in the intervention group who visited the practice saw just the practice nurse.
Scenario 2. . If women were invited at the same time as they received their result letter to consult the nurse, the only extra nurse time spent on the intervention group compared with the control group would be that resulting from the longer consultation time (difference between means of 8.7 minutes).
Scenario 3. . If use of the packages became policy, the associated costs would be spread over a large number of consultations. During the 14 months of the trial, the smear tests of 607 women from 125 practices were reported as mildly dyskaryotic. Although not all of these women were eligible to take part in the trial (including 34 no longer registered with participating practices when they received their smear test result), this is equivalent to a potential of 4.2 women per practice per year. The package was estimated to remain in use for 5 years: this allows for a realistic physical and informative life, in addition to possible replacement by contemporary media, for example, computer applications. The one-off costs are therefore shared over 21 consultations in Scenario 3. Scenario 3 also assesses the effects of high-volume production of the package (relevant to a health authority) and low-volume production (for an individual practice).
Multi-scenario analysis combines all three scenarios to calculate the minimum anticipated cost in practice. Mean NHS and practice costs per woman are then reported in a format relevant to decision makers. These are reported for a range of (i) practice sizes and (ii) screening areas, using the costs per woman from the multi-scenario analysis.
|
Results |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionSummary: key points for...References |
|---|
There are two distinct elements of costs: the cost of the intervention package and training, and the cost of consultations subsequent to receipt of a mildly abnormal smear test result.
Package and training costs
The actual cost of printing each educational package for the trial was £9.48. Subsequent quotations ranged from £8.17 each for 100 or more, to £10.90 for single prints.
Thirty-one (65%) post-consultation questionnaires were returned. The majority (93%) of training sessions took place in practice time. The mean time spent in training sessions was 22 minutes per practice. This has been valued at the cost per minute (£0.27) of nurses' working time.6 The mean cost to a practice of a training session within the trial is therefore £5.94.
The package and training involve a total one-off cost of £15.42 (£9.48 + £5.94) per practice within the trial. A mean of 2.35 women were recruited per trained intervention practice. Therefore, the package and training costs incurred per woman in the intervention arm of the trial (whether or not she attended) were £15.42/2.35, that is, £6.56. This figure is for the number of women recruited to the trial; outside the confines of the trial, these costs could be shared over a much larger number of women.
Consultation costs
Almost twice as many women in the intervention group visited their practice to discuss their result with one or more health professionals (73% of 113 compared with 37% of 127 controls). Moreover, in the intervention group, 66 (62%) women saw the practice nurse and 23 (22%) saw the GP; equivalent figures for the control group were seven (6%) and 39 (31%), respectively.
Nurses in the intervention group spent on average an extra 8.7 minutes in face-to-face contact with women (19.3 minutes in the intervention group and 10.6 minutes in the control group). Intervention nurses also spent an additional 14.3 minutes in preparation for consultations, primarily telephoning/writing to women to invite them to make an appointment. Women from the control group who attended the practice made their own appointments in the usual way through reception. This process would not have involved any nurse time and any additional cost of reception time has not been included. Therefore, nurses in the intervention practices spent, on average, an extra 23 minutes per consultation compared with nurses in the control practices.
The mean total time (including waiting) per woman spent at the practice when consulting a GP was 30 and 33 minutes for the intervention and control groups, respectively. As there was little difference between the two groups, a standard length of time for a GP consultation of 9.3 minutes6 is assumed for both arms of the trial.
Total NHS and practice costs
Table 1
shows costs to the NHS and practices in terms of GP and nurse time per consultation. A practice is reimbursed by the health authority for 70% of the cost of employing a practice nurse; therefore only 30% of nurse costs are included for the analysis from the practice perspective. Table 2 gives the mean total costs per woman.
|
|
Sensitivity analysis
Table 3
gives the results of the sensitivity analysis alongside the control group costs and the observed trial costs—the latter can be viewed as likely maximum costs of the intervention.
|
Altering the projected pattern of visits in accordance with Scenario 1 (all consultations with nurses) reduces both the NHS costs and the practice costs because nurse consultations incur a lower cost than GP consultations. The reduction is greater for the practices since the nurse costs are shared as described above. In neither case are the costs reduced to the levels in the control group since the costs of the package remain relatively high.
As expected in Scenario 2 (elimination of pre-consultation preparation time), costs are reduced more from the NHS perspective, which incurs 100% of nurse costs, but again they remain higher than the control group.
By allocating the package and training costs over a large number of consultations (Scenario 3), the intervention practice projected costs approach those of the control practices, although they remain higher for the NHS as a whole. Varying the production volume has almost no effect on the costs.
As an indication of the minimum anticipated costs of introducing the intervention, the multi-scenario analysis shows that practice costs are lower for intervention than control practices, and that the additional costs to the NHS are £1.40 per woman in the intervention compared with the control group (Table 3
).
Costs for screening areas and individual practices
Table 4 shows the NHS costs over 1 year for a health authority with a screening service carrying out 60 000 or 100 000 tests a year, with either 1 or 3% of smear tests returning a mildly dyskaryotic result. Table 5 gives costs together with current and expected numbers of consultations with GPs and practice nurses for different sizes of practice (in terms of number of partners).
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionSummary: key points for...References |
|---|
From the perspective of the practices, a change from current policy to the intervention policy led to potential (negligible) savings of around £3.50 per partner per year. From the NHS perspective, the intervention would lead to slightly increased costs of between £1000 and £2500 per year for an area performing 60 000 tests per year. Both these findings are based on the assumptions within the multi-scenario sensitivity analysis and are relevant to any cervical screening programme that is centrally organized, publicly funded and executed through the primary care setting.
Women in the intervention group were invited, and therefore more likely, to consult the practice nurse than the GP. This increased the total number of consultations as well as shifting some care from the GP to the nurse. Nurse consultations were longer than GP consultations. The reduction in costs from the practice perspective is a result of this shift in care to the nurse, whose costs are shared between the practice and the health authority. The increased costs to the health authority of this same shift are due to a combination of factors: the authority being responsible for the whole cost of the nurse, the rise in the number of consultations and the increased length of consultations.
Data on consultations to discuss the smear test result were obtained from questionnaires completed by the women. Although there is potential for under- or over-reporting (which may not be consistent across the randomization groups), this approach had the advantage that any consultations during which the smear test result was discussed were included. This would not necessarily have been possible if this information had been extracted from patient records.
In order to enable decision makers to extrapolate these results to their own setting,9 the cost per consultation with both GP and practice nurse under the trial conditions is given in Table 1, as well as the projected cost per woman under various scenarios (Table 3). These figures can be adapted to a range of settings using local data about numbers of women under surveillance and attendance rates. For example, the health authority areas in which this trial was undertaken have a policy of screening women once every 5 years. In areas where the policy is 3-yearly screening, the mean number of mildly dyskaryotic results per practice may be greater than those used in Tables 4 and 5. Similarly, women with smear results showing borderline changes are usually placed under surveillance. The inclusion of these women in those offered the intervention would also increase the figures given in Tables 4 and 5.
The main implication for general practices of the introduction of the policy evaluated in the trial is the change in the pattern of care provided: fewer women consulted their GP about their smear result and many more, following active encouragement, consulted the practice nurse. The appropriateness and feasibility of such a shift in care would be expected to be influenced by the particular circumstances prevailing in individual practices. One further consideration is the possibility that the intervention might have influenced the attendance rates for the repeat smear test. In the event, however, 85% of women attended in the intervention group compared with 83% in the control group. Had substantially different proportions attended in the two groups, the economic consequences and implications for long-term health of the women could be considerable and would have to be accounted for in the evaluation.
The costs of delivering the intervention depend on a number of assumptions. The actual costs of incorporating the package into routine clinical practice will undoubtedly be different to the trial costs. In order to increase the external validity of the costs,7,10 the sensitivity analysis involved realistic alterations to the way in which the package could be delivered. The resultant minimum anticipated costs from the multi-scenario analysis were used to derive Tables 4 and 5 to reflect a best estimate of actual costs in practice.
There are a number of options for timing and selection of components of the package, including tailoring its delivery to individual women's requirements or adding further information.11 Indeed, some information from the package could be given when the smear test is performed. Alternatively, the package could be used opportunistically in primary care.
Although the effectiveness data showed few differences in anxiety between the intervention and control women,5 significantly more women in the intervention group were content to wait 6 months for their repeat smear test. The cost implications presented in this paper should be borne in mind when considering whether to implement such an educational intervention in order to derive this benefit.
|
Summary: key points for the primary care team |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionSummary: key points for...References |
|---|
- It is both feasible and acceptable for practice nurses to deliver the educational intervention and this reduces the number of GP consultations.
- From the perspective of the NHS, the overall cost of implementing the package is between £1000 and £2500 for an area performing 60 000 tests per year.
- From the perspective of a practice, the policy is effectively cost-neutral.
|
Acknowledgments |
|---|
We thank Irene Jones and Mags Rees for their major contribution in the conducting of this study, and to Caroline Joy, Alun Evans and Elizabeth McKenzie from the two Screening Units covered by the trial. We are very grateful to the women who took part in the trial, and to the participating practices, especially the practice nurses.
Funding for this research was provided by the Cancer Research Campaign and the South & West NHS R&D Directorate.
The Division of Primary Care and the Department of Social Medicine at the University of Bristol, and the Division of General Practice at the University of Wales College of Medicine, are part of the Medical Research Council Health Services Research Collaboration.
|
References |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionSummary: key points for...References |
|---|
1 Morris M, Tortolero-Luna G, Malpica A et al. Cervical intraepithelial neoplasia and cervical cancer. Obstet Gynecol Clin North Am 1996; 23: 347–410.[Web of Science][Medline]
2 Campion MJ, Brown JR, McCance DJ, Atia W, Edwards R. Psychosexual trauma of an abnormal cervical smear. Br J Obstet Gynaecol 1988; 95: 175–181.[Web of Science][Medline]
3 Lerman C, Miller S, Scarborough R, Hanjani P, Nolte S, Smith D. Adverse psychologic consequences of positive cytologic cervical screening. Am J Obstet Gynecol 1991; 165: 658–661.[Web of Science][Medline]
4 Bell JS, Kitchener HC, Fisher PM, Mann E. Psychological Response to Cervical Screening. Report K/OPR/2/2/C891 to Health Services Research Committee, Home and Health Department, Scottish Office.
5 Peters TJ, Somerset M, Baxter K, Wilkinson C. Anxiety amongst women with mild dyskaryosis: a randomised trial of an educational intervention. Br J Gen Pract 1999; 49: 348–352.[Medline]
6 Netten A, Dennett J. Unit Costs of Health and Social Care 1996. University of Kent: Personal Social Services Research Unit, 1996.
7 Briggs AH, Sculpher MJ, Buxton MJ. Uncertainty in the economic evaluation of healthcare technologies: the role of sensitivity analysis. Health Economics 1994; 3: 95–104.[Web of Science][Medline]
8
Drummond MF, Jefferson TO. Guidelines for authors and peer reviewers of economic submissions to the BMJ. Br Med J 1996; 313: 275–283.
9 Bryan S, Brown J. Extrapolation of Cost-effectiveness Information to Local Settings. HERG Discussion Paper No. 17. Brunel University: Health Economists Research Group, 1997.
10 Drummond MF, Stoddart GL, Torrance GW. Methods for the Economic Evaluation of Health Care Programmes. 2nd edn. Oxford: Oxford University Press, 1997.
11 Somerset M, Peters T. Intervening to reduce anxiety: do we know what works for women with mild dyskaryosis? J Adv Nurs 1998; 28(3): 563–570.[Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||