Family Practice Vol. 18, No. 1, 1-8
© Oxford University Press 2001
Research in Practice |
A randomized controlled trial of direct access for laparoscopic sterilization
Dugald Baird Centre for Research on Women's Health,
a Department of General Practice and Primary Care,
b Health Economics Research Unit and
d Department of Obstetrics and Gynaecology, University of Aberdeen and
c Grampian University Hospitals NHS Trust, UK.
Correspondence to Dr Blair H Smith, Department of General Practice and Primary Care, University of Aberdeen, Foresterhill Health Centre, Westburn Road, Aberdeen AB25 2AY, UK.
McKessock L, Smith BH, Scott A, Graham W, Terry PB, Templeton A and Fitzmaurice AE. A randomized controlled trial of direct access for laparoscopic sterilization. Family Practice 2001; 18: 1–8.
Received 2 February 2000; Revised 13 June 2000; Accepted 5 September 2000.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionAppendix 1 Referral proformaAppendix 2 Example of...References |
|---|
Objectives. The aim of this study was to establish and evaluate a direct access service for laparoscopic sterilization.
Methods. A pragmatic randomized controlled trial was carried out on a total of 232 women referred to Aberdeen Royal Infirmary for sterilization between 1 June 1996 and 31 March 1997, from 57 general practices around Aberdeen, Scotland, comprising 75 from general practices that had received criteria for direct access, and 157 from control practices. The main outcome measures were: patients' waiting times to appointments and operation, satisfaction, short-term regret, operative complications and costs; conjoint analysis of patient preferences; GPs' adherence to referral criteria; GP and gynaecologist satisfaction; and NHS costs.
Results. Analysis was by intention to treat. There were no inappropriate direct referrals. Waiting time was lower in the intervention group, but there were more visits post-operatively to the GP. Patient and doctor satisfaction was equally high in both groups. There was no difference in operative complication rate, nor in total cost to patients or to the NHS. GPs and gynaecologists strongly supported direct access, but women preferred to meet both a gynaecologist and a GP before sterilization (routine referral). Other attributes of care important to patients included written information, although waiting time was not important. Of all women referred for sterilization during the study period, only 31% were suitable for direct referral.
Conclusions. Patients preferred routine referral, and there were no differences in patient costs or satisfaction. There was, however, strong medical support for direct referral. Restrictive direct referral criteria may limit the uptake. These findings are important for future planning of direct referral services. Important methodological lessons were also learnt about the conduct of trials involving the primary–secondary care interface.
Keywords. Conjoint analysis, direct access, direct referral, laparoscopic sterilization, randomized controlled trial.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAppendix 1 Referral proformaAppendix 2 Example of...References |
|---|
The process and outcome of referral between GP and specialist for many routine procedures can be inefficient. It may involve duplication of skills and impose increased pressures on out-patient clinics.1,2 Laparoscopic sterilization is an area which might be rendered more efficient and satisfactory to patients and providers by implementing a direct referral service with referral guidelines and patient advice leaflets. Each year, ~500 laparoscopic sterilizations are performed in Aberdeen Royal Hospitals NHS Trust. We report the results of a randomized controlled trial (RCT) of laparoscopic sterilization from general practice, comparing routine referral through the Gynaecology Out-patient Department (GOPD) versus direct access. Our hypothesis was that a direct access service for laparoscopic sterilization could be run with equal or greater patient satisfaction and safety compared with routine referral, but at less cost to the NHS and with greater satisfaction for GPs and gynaecologists.
RCTs are argued to be the ‘gold standard’ for assessing new models of care and comparing them with existing models.3 RCTs present considerable methodological challenges, however, particularly where a variety of parties and outcome measures are involved. As well as learning from the conclusions of the trial, there are often methodological lessons to be learnt, and it is important to report both. Although the objective of this study was to establish and evaluate a new referral service, we will take this opportunity also to report on some methodological issues raised.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAppendix 1 Referral proformaAppendix 2 Example of...References |
|---|
Study population
All 72 general practices in Grampian, Scotland who referred patients to Aberdeen Royal Infirmary (ARI) for sterilization in 1996 were invited to participate. Details of the study's aims were issued with a clear indication of the level of involvement required. A total of 61 (85%) practices agreed to participate. These practices were randomized into intervention and control groups, stratified by list size, fundholding status and rural or urban location.
Subjects
From June 1996 to March 1997, a total of 283 women were referred for sterilization to ARI. Of these, 232 women aged 22–46 years were referred from the 57 practices participating in the study. Seventy-five women were referred from intervention practices, of whom only 14 were eligible for direct referral using the study criteria. Using the routine referral system through the GOPD, 157 women were referred by those practices in the control arm (see Fig. 1
).
|
Women's consent to participate in the study was obtained by the referring GPs. Women who declined participation were referred by the usual route, their care was unaffected and they were excluded from the study. Upon admission to hospital via the direct referral service, women discussed the procedure with the consultant or senior registrar who would undertake the operation to confirm adequacy of pre-operative counselling. Theatre lists of two consultants were dedicated to direct referral sterilizations.
Referral criteria
The referral criteria were drawn up by a multidisciplinary team including gynaecologists, GPs and health service researchers. There was widespread consultation with local gynaecologists and GPs before the referral criteria were finalized. These were based on the knowledge and expert opinion of those involved and evidence of risk factors for regret following sterilization from the literature,4 and aimed to ensure a safe and uncomplicated sterilization.
The criteria were supplied to GPs as part of a referral pack, which included a detailed referral sheet, a structured referral proforma concerning relevant patient history, examination details and counselling provided (Appendix 1). The referral proforma was sent directly to the research nurse. A patient information booklet Laparoscopic Sterilization—Your Questions Answered, adapted from a document produced by the Royal College of Obstetricians and Gynaecologists, was issued to women in the intervention arm of the study at the time of GP consultation.
Interventions
Practices in the intervention group referred women for laparoscopic sterilization using the referral criteria (see later) developed for the trial. A newsletter kept practices up to date with the study's progress and encouraged continuing participation.
Outcome measures
The primary outcome measures were patient satisfaction, operative complication rate, and patient and NHS costs incurred. The secondary outcome measures were waiting times from referral to operation, short-term regret, GPs' adherence to direct referral criteria, and GP and gynaecologist satisfaction. There was also a conjoint analysis of patients' preferences.
The following research instruments were designed, piloted and used in the study:
- (i) pre-admission questionnaire: this assessed the women's knowledge of the procedure to ensure adequate pre-operative counselling;
- (ii) patient satisfaction and costs questionnaire: this was delivered to women by the research sister during their admission for sterilization;
- (iii) two-month follow-up questionnaire: this postal instrument included: (a) patient satisfaction and short term-regret 2 months post-operatively; and (b) conjoint analysis of patients' preferences; and
- (iv) GP and gynaecologist satisfaction questionnaire.
- (ii) patient satisfaction and costs questionnaire: this was delivered to women by the research sister during their admission for sterilization;
All women referred for sterilization to ARI during the study period from control and intervention practices were assessed by review of their hospital records for suitability for direct referral, according to the referral criteria. Hospital records were also examined for waiting times between referral and operation, and for intra- and post-operative complications.
Sample size
For calculation of the required sample size, we employed the method used by Makuch and Simon, cited in Pocock5 for ‘negative’ trials (i.e. "trials in which one is more interested in showing a ‘negative’ result namely that the two treatments are equally effective"). We hypothesized that routine and direct referral would result in equal patient satisfaction, with reduced costs being the potential primary benefit from direct referral. Assuming that 80% of patients are satisfied with the current referral system,6 we calculated that with a total sample size of 253 women we could demonstrate with 95% confidence that this figure after direct referral would be at worst 70% and at best 90%, with a power of 80%. This sample size represented approximately one half of all laparoscopic sterilizations performed at ARI in the previous year, and this was felt to be achievable within the project time frame.
Analysis
Data were analysed using the SPSS for Windows software package. The principle of intention to treat was adopted for the primary analysis using the chi-square test, and t-tests, Fisher's exact test and the difference between two proportions as appropriate.
Economic analysis
The study tracked NHS resources used by participating patients and included all GP and hospital contacts until 2 months after the patient's operation. The cost of out-patient visits and the operation were included for patients who did not attend, as this still incurs an opportunity cost. Information about the quantities of resources used (e.g. number of visits) was obtained from the patient satisfaction and follow-up questionnaires. Costs were in 1996 prices and were based on information from Netten and Dennett,7 the Finance Department at Aberdeen Royal Hospitals NHS Trust and a previous detailed costing of a laparoscopic sterilization. Unit costs used are available from the authors. Patients' time costs (and those of relatives who accompanied them) were calculated using mean wage rates. Patient's out-of-pocket costs were also included.
Patients' preferences for a range of hypothetical scenarios describing different types of care for laparoscopic sterilization were elicited using conjoint analysis.8,9 Individuals in the intervention and control group were asked to choose between a series of pairs of these scenarios (A or B) in a self-administered questionnaire (see example in Appendix 2). Scenarios differed according to four attributes: (i) the amount of written information provided (no information, a simple leaflet or a more detailed booklet); (ii) waiting time to operation (2, 4 and 6 months); (iii) who is consulted before the operation (your GP, a hospital doctor or both); and (iv) travel costs (£15, £30 or £45). These data were analysed using random effects probit regression with the respondent's choice of either scenario A or B as the dependent variable. As well as including the main attributes, interaction terms were also created to see if the relative importance of main attributes depended on age, marital status, employment status, income, education, direct referral and whether the importance of waiting time depended on their actual waiting time. All variables were entered into the model and then excluded one at a time using a P-value of 
0.10. Further details are available from the authors.
Assignment
Although sample size calculation was based on number of patients, reflecting the most important outcome measures, randomization was by referring practice to encourage consistency of referral behaviour and to avoid administrative complications within practices. Participating practices were randomized into intervention and control groups, stratified by list size, fundholding status and rural or urban location. The practices in the eight blocks were placed into different coloured envelopes (one colour for each block). A coin was flipped to decide to commence with intervention or control allocation. From the first block of small list practices an envelope was selected and at the same time an envelope from a large list block was selected. These practices were randomized to the same arm of the trial and this process continued, allocating control or intervention alternately until all practices in those two groups were allocated into one of the two arms of the trial: one receiving and implementing the guidelines (intervention arm) and one maintaining the status quo (control arm). This process was carried out for each of the eight blocks.
Masking
In view of the nature of the intervention, it was not possible to mask patients or practices. Practicalities made masking of gynaecologists difficult, and in any case this was not safe as adequacy of pre-operative counselling and appropriateness of direct referral had to be assured by operators prior to operation.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAppendix 1 Referral proformaAppendix 2 Example of...References |
|---|
Participant flow
Following randomization, four of the 61 practices withdrew from the study, all from the intervention arm, citing the following reasons: new head of practice; increased workload; and a perceived inappropriate shift in workload (two practices). This left 57 practices (79% of the Grampian total) participating in the study (31 in the control arm and 26 in the intervention), encompassing 230 GPs.
The study was terminated before the expected date, after discussion with the funding body, as lower than expected recruitment rates made timely completion impossible. The findings of the trial are reported here despite the small numbers since important lessons have been learnt. During the 10 month data collection period, recruitment ceased with 75 women from intervention practices and 157 from control practices, which is just over half of the expected number for that period. Progress of all women referred for sterilization during the study period is shown in Figure 1
.
In terms of overall satisfaction, there were no significant differences between the two study groups, with 89% of the control group (89/100) and 93% of the intervention group (43/46) expressing overall satisfaction with the service provided. Those women in the intervention group found the pre-operative information more helpful compared with those in the control group (P = 0.003). There was no difference in satisfaction with subsequent hospital stay between the two patient groups. The generally high level of satisfaction was maintained, with no differences between groups at 2 months post-operatively.
After review of the hospital records of all women referred from participating practices, fewer than one-third were found to be suitable for direct referral according to the criteria (Table 1). There were no significant differences found in patients' characteristics between the control and intervention groups. Data extracted from operation notes revealed no significant differences in rates of intra-operative complications.
|
Women in the intervention group were found to have shorter waiting times from referral to operation. However, responses to the questionnaires found that they spent longer with their GP before the operation and had more visits afterwards (Table 2
). Patients' time and travel costs were not significantly different between intervention and control groups (Table 3).
|
|
Table 4
shows the coefficients from the final reduced regression model from the conjoint analysis. Variables that were not significant were removed from the model. The size of the coefficients represents their relative importance. The travel cost attribute did not influence which scenario was chosen (i.e. it was not important to patients). Waiting time was not an important factor by itself. The most important attribute (given by the size of the coefficient) was the amount of written information provided. Respondents' income, education, marital status, employment status and satisfaction with their GP visit did not influence their preferences. The direction of preference for the levels of each attribute was established using the coefficients in Table 4 and the categorical codes assigned to the levels of each attribute.
|
Respondents showed a stronger preference for seeing their GP and hospital doctor to discuss sterilization, rather than seeing only their GP. This tends to support the current system of referral. Women preferred more information to less. Older respondents preferred shorter waiting times and required less written information than younger respondents.
Doctors' opinions of the referral service and guidelines were elicited from questionnaires sent to all GPs in participating practices. The response rate was 74% (171/230). Questionnaires were also sent to all consultants and specialist registrars at ARI, which yielded a response rate of 85% (22/26). There was a high level of support among GPs and gynaecologists for direct referral, 85 and 77% approving, respectively. The criteria were accepted by most as appropriate; 189/230 GPs and 18/26 gynaecologists. However, 30/230 GPs and 8/26 gynaecologists felt the criteria were too restrictive. None of the differences between GPs' and gynaecologists' responses were statistically significant.
GP costs were significantly higher in the intervention group (caused by increased number of visits after the operation). Total costs to patients and total NHS cost per patient were not significantly different in either group. Sensitivity analysis examined different methods of calculating GP and hospital costs. If the tariff for a GOPD referral was used, rather than a cost per visit, then the difference in total NHS cost approached statistical significance at the 5% level. However, the latter method reflects the opportunity costs more accurately.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAppendix 1 Referral proformaAppendix 2 Example of...References |
|---|
To the authors' knowledge, this is the first published RCT of direct access for laparoscopic sterilization. The findings will be of interest to those considering establishment of a direct referral service. In proposing this research, we suggested that if patient satisfaction was no worse but the cost of direct referral was less than routine referral, then a more efficient service could be introduced. This study, however, found no difference in patient satisfaction or total cost between the two methods of referral. The only reason found in this study to support a direct referral service was that GPs and gynaecologists favoured it, and this approval, though high, was not unanimous and possibly was biased by withdrawal and non-inclusion of GPs opposed to this type of service. These findings indicate the need to conduct studies such as this before implementing new services based on similar assumptions. The small sample size achieved may have contributed to the absence of statistically significant findings for costs and patient satisfaction.
The results of the conjoint analysis, however, were statistically robust, and highlight the relative importance of good quality patient literature relative to waiting times. Although women were equally satisfied with their experiences, if given the choice, women would prefer the status quo. This is likely to reflect the fact that women preferred what they had experienced. Older women were less concerned about the loss of a GOPD consultation, however, and the referral criteria, allowing only women over the age of 30 to be referred directly, may be appropriate for older groups.
The relative unimportance of waiting times may reflect the fact that waiting longer for sterilization does not worsen health or quality of life, as would be the case for acute, symptomatic or deteriorating conditions. Women were more concerned about talking to their GP and gynaecologist and obtaining good written information about sterilization.
There were low rates of post-operative complications and short-term regret amongst women in the treatment group. There were no inappropriate direct referrals. Concerns about higher rates of inappropriate referral, post-operative complications and post-sterilization regret were not realized in this study. Further studies of long-term regret are required. It is important to note the relatively small number of women who would be suitable for direct referral for laparoscopic sterilization (31%), despite general acceptance of the appropriateness of direct referral criteria.
Finally, there is a suggestion emerging from this study that the rate of female sterilization request itself is falling in Grampian, given the shortfall between predicted referral rate (based on the previous year's figures) and the actual rates. This has important clinical and reproductive (or contraceptive) implications. It may be that female sterilization is being superseded as a method of contraception by vasectomy (regarded as safer) or innovative techniques such as the Mirena® intra-uterine system, and this warrants further investigation.
Lessons learnt
First, the referral criteria were designed to make the service safe and feasible, but were therefore restrictive. Data collected on all women referred during the study period suggested that only one-third (31%) of women in Grampian were eligible for direct referral according to the study criteria. Any future research on direct referral must allow for this small proportion.
Secondly, difficulties were encountered in persuading many GPs to participate in the study. Without clear incentives, it was sometimes difficult for GPs to agree to undertake more work, even with the hope of gaining a more efficient service. Withdrawals after randomization led to an uneven balance between control and intervention groups, which compromised the study. Consideration might be given to financial remuneration for GPs who participate in trials, in recognition of the time involved and consistent with their role as business operators.10 This might maximize the success of any trial but would add considerably to the expense, and represent a notable precedent.
Thirdly, lessons were also learnt regarding sample size determination. Underlying changes in the demand for the procedure were not predicted but affected recruitment. It appeared that in Grampian, the rate of sterilization request was decreasing. Routine recent retrospective data may not give an accurate picture of demand, and further research needs to be undertaken to predict demand before trials are begun. Consideration should also be given to sample size calculations based on economic outcomes.
|
Appendix 1 Referral proforma |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAppendix 1 Referral proformaAppendix 2 Example of...References |
|---|
|
|
Appendix 2 Example of scenario |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAppendix 1 Referral proformaAppendix 2 Example of...References |
|---|
|
|
Acknowledgments |
|---|
We are very grateful to all the women, GPs and gynaecologists who contributed to this study for their time and commitment. This study was funded by a project grant from the Chief Scientist Office, Scottish Office Home and Health Department (SOHHD). The Health Economics Research Unit is also funded by the same office. The views expressed are those of the authors and not SOHHD.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAppendix 1 Referral proformaAppendix 2 Example of...References |
|---|
1 Smith FC, Gwynn BR. Direct access surgery. Ann R Coll Surg Engl 1995; 77: 94–96.[Medline]
2 Bradshaw C, Pritchett CJ, Wright H, Eccles M, Armitage T, Todd E. The South Tyneside Fastrak Service—background and development. J One Day Surg 1994; Spring: 628.
3
Chalmers I. Unbiased, relevant, and reliable assessment in health care. Br Med J 1998; 317: 1167–1168.
4 Wilcox LS, Chu SY, Eaker ED, Zeger SL, Peterson HB. Risk factors for regret after tubal sterilisation: 5 years of follow-up in a prospective study. Fertil Steril 1991; 55: 927–933.[Web of Science][Medline]
5 Pocock SJ. Clinical Trials: A Practical Approach. Chichester: Wiley, 1983.
6 The Audit Commission for Local Authorities and the National Health Service in England and Wales. Measuring Quality: The Patient's View of Day Surgery. London: HMSO, 1991.
7 Netten A, Dennett J. Unit Costs of Health and Social Care. Canterbury: PSSRU, University of Kent at Canterbury, 1996.
8 Ryan M. Using Consumer Preferences in Health Care Decision Making. The Application of Conjoint Analysis. London: Office of Health Economics, 1996.
9 Vick S, Scott A. Agency in health care. Examining patients' preferences for attributes of the doctor–patient relationship. J Health Econ 1998; 17: 587–606.[Web of Science][Medline]
10
Foy R, Parry J, McAvoy B. Clinical trials in primary care. Br Med J 1998; 317: 1168–1169.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||