The clinical effectiveness and cost-effectiveness of screening for open angle glaucoma: a systematic review and economic evaluation

OBJECTIVES

To assess whether open angle glaucoma (OAG) screening meets the UK National Screening Committee criteria, to compare screening strategies with case finding, to estimate test parameters, to model estimates of cost and cost-effectiveness, and to identify areas for future research.

DATA SOURCES

Major electronic databases were searched up to December 2005.

REVIEW METHODS

Screening strategies were developed by wide consultation. Markov submodels were developed to represent screening strategies. Parameter estimates were determined by systematic reviews of epidemiology, economic evaluations of screening, and effectiveness (test accuracy, screening and treatment). Tailored highly sensitive electronic searches were undertaken.

RESULTS

Most potential screening tests reviewed had an estimated specificity of 85% or higher. No test was clearly most accurate, with only a few, heterogeneous studies for each test. No randomised controlled trials (RCTs) of screening were identified. Based on two treatment RCTs, early treatment reduces the risk of progression. Extrapolating from this, and assuming accelerated progression with advancing disease severity, without treatment the mean time to blindness in at least one eye was approximately 23 years, compared to 35 years with treatment. Prevalence would have to be about 3-4% in 40 year olds with a screening interval of 10 years to approach cost-effectiveness. It is predicted that screening might be cost-effective in a 50-year-old cohort at a prevalence of 4% with a 10-year screening interval. General population screening at any age, thus, appears not to be cost-effective. Selective screening of groups with higher prevalence (family history, black ethnicity) might be worthwhile, although this would only cover 6% of the population. Extension to include other at-risk cohorts (e.g. myopia and diabetes) would include 37% of the general population, but the prevalence is then too low for screening to be considered cost-effective. Screening using a test with initial automated classification followed by assessment by a specialised optometrist, for test positives, was more cost-effective than initial specialised optometric assessment. The cost-effectiveness of the screening programme was highly sensitive to the perspective on costs (NHS or societal). In the base-case model, the NHS costs of visual impairment were estimated as 669 pounds. If annual societal costs were 8800 pounds, then screening might be considered cost-effective for a 40-year-old cohort with 1% OAG prevalence assuming a willingness to pay of 30,000 pounds per quality-adjusted life-year. Of lesser importance were changes to estimates of attendance for sight tests, incidence of OAG, rate of progression and utility values for each stage of OAG severity. Cost-effectiveness was not particularly sensitive to the accuracy of screening tests within the ranges observed. However, a highly specific test is required to reduce large numbers of false-positive referrals. The findings that population screening is unlikely to be cost-effective are based on an economic model whose parameter estimates have considerable uncertainty. In particular, if rate of progression and/or costs of visual impairment are higher than estimated then screening could be cost-effective.

CONCLUSIONS

While population screening is not cost-effective, the targeted screening of high-risk groups may be. Procedures for identifying those at risk, for quality assuring the programme, as well as adequate service provision for those screened positive would all be needed. Glaucoma detection can be improved by increasing attendance for eye examination, and improving the performance of current testing by either refining practice or adding in a technology-based first assessment, the latter being the more cost-effective option. This has implications for any future organisational changes in community eye-care services. Further research should aim to develop and provide quality data to populate the economic model, by conducting a feasibility study of interventions to improve detection, by obtaining further data on costs of blindness, risk of progression and health outcomes, and by conducting an RCT of interventions to improve the uptake of glaucoma testing.

A randomised controlled trial of postoperative radiotherapy following breast-conserving surgery in a minimum-risk older population. The PRIME trial

OBJECTIVES

To assess whether omission of postoperative radiotherapy in women with 'low-risk' axillary node negative breast cancer (T0-2) treated by breast-conserving surgery and endocrine therapy improves quality of life and is more cost-effective.

DESIGN

A randomised controlled clinical trial, using a method of minimisation balanced by centre, grade of cancer, age, lymphatic/vascular invasion and preoperative endocrine therapy, was performed. A non-randomised cohort was also recruited, in order to complete a comprehensive cohort study.

SETTING

The setting was breast cancer clinics in cancer centres in the UK.

PARTICIPANTS

Patients aged 65 years or more were eligible provided that their cancers were considered to be at low risk of local recurrence, were suitable for breast-conservation surgery, were receiving endocrine therapy and were able and willing to give informed consent.

INTERVENTIONS

The standard treatment of postoperative breast irradiation or the omission of radiotherapy.

MAIN OUTCOME MEASURES

Quality of life was the primary outcome measure, together with anxiety and depression and cost-effectiveness. Secondary outcome measures were recurrence rates, functional status, treatment-related morbidity and cosmesis. The principal method of data collection was by questionnaire, completed at home with a research nurse at four times over 15 months.

RESULTS

The hypothesised improvement in overall quality of life with the omission of radiotherapy was not seen in the EuroQol assessment or in the functionality and symptoms summary domains of the European Organisation for Research in the Treatment of Cancer (EORTC) scales. Some differences were apparent within subscales of the EORTC questionnaires, and insights into the impact of treatment were also provided by the qualitative data obtained by open-ended questions. Differences were most apparent shortly after the time of completion of radiotherapy. Radiotherapy was then associated with increased breast symptoms and with greater fatigue but with less insomnia and endocrine side-effects. Patients had significant concerns about the delivery of radiotherapy services, such as transport, accommodation and travel costs associated with receiving radiotherapy. By the end of follow-up, patients receiving radiotherapy were expressing less anxiety about recurrence than those who had not received radiotherapy. Functionality was not greatly affected by treatment. Within the randomised controlled trial, the Barthel Index demonstrated a small but significant fall in functionality with radiotherapy compared with the no radiotherapy arm of the trial. Results from the non-randomised patients did not confirm this effect, however. Cosmetic results were better in those not receiving radiotherapy but this did not appear to be an important issue to the patients. The use of home-based assessments by a research nurse proved to be an effective way of obtaining high-quality data. Costs to the NHS associated with postoperative radiotherapy were calculated to be of the order of 2000 pounds per patient. In the follow-up in this study, there were no recurrences, and the quality of life utilities from EuroQol were almost identical.

CONCLUSIONS

Although there are no differences in overall quality of life scores between the patients treated with and without radiotherapy, there are several dimensions that exhibit significant advantage to the omission of irradiation. Over the first 15 months, radiotherapy for this population is not a cost-effective treatment. However, the early postoperative outcome does not give a complete answer and the eventual cost-effectiveness will only become clear after long-term follow-up. Extrapolations from these data suggest that radiotherapy may not be a cost-effective treatment unless it results in a recurrence rate that is at least 5% lower in absolute terms than those treated without radiotherapy. Further research is needed into a number of areas including the long-term aspects of quality of life, clinical outcomes, costs and consequences of omitting radiotherapy.

Geographical access to imaging facilities for stroke patients in Scotland

This study examines the geographical access to imaging facilities for suspected stroke patients in Scotland. A survey of Scottish clinical directors of radiology was initially undertaken to determine the current and future provision of brain imaging for the diagnosis of stroke. We analysed geographical and digital population data with geographical information systems software to determine access to brain imaging services for stroke patients during 'normal' working hours and 'out-of-hours'. The findings suggest that, in general, most departments are able to deliver scanning for stroke as set within current guidelines, at least in normal working hours. However, radiological departments are generally operating at full capacity, and there is restricted availability of scanning services for stroke in certain regions during weekend periods. It is vital that policy makers consider these findings when reviewing the guidelines for recommending scanning for stroke.

Screening for abdominal aortic aneurysm in a geographically isolated area

Background

Screening for abdominal aortic aneurysm has been shown to reduce aneurysm-related mortality, but the applicability of the results to the whole of the UK has been questioned. This study examined screening in a remote and rural area.

Methods

Over 3 years, men aged 65–74 years were offered screening in the community by ultrasonography, usually in general practitioner surgeries. Men with an aneurysm were rescanned at intervals or assessed for surgery. The screening and hospital costs of the programme were calculated.

Results

Some 9323 men were offered screening of whom 8355 (89·6 per cent) attended. Uptake was high in all areas. A total of 430 scans (5·1 per cent) were abnormal; 40 men had an aneurysm greater than 55 mm in diameter. Twenty further men had an aorta that enlarged to greater than 55 mm during follow-up. A total of 54 men had elective repair with one death (mortality rate 2 per cent). The cost of screening alone was £16 per invitation and the overall cost of the programme, including surgery, was £58 per invitation.

Conclusion

Screening for abdominal aortic aneurysm can be carried out in a remote and rural area with high uptake, acceptable clinical results and at no greater cost than in more densely populated areas.

What is the best imaging strategy for acute stroke?

OBJECTIVES

To determine the cost-effectiveness of computed tomographic (CT) scanning after acute stroke. To assess the contribution of brain imaging to the diagnosis and management of stroke, and to estimate the costs, benefits and risks of different imaging strategies in order to provide data to inform national and local policy on the use of brain imaging in stroke.

DESIGN

A decision-analysis model was developed to represent the pathway of care in acute stroke using 'scan all patients within 48 hours' as the comparator against which to cost 12 alternative scan strategies.

SETTING

Hospitals in Scotland.

PARTICIPANTS

Subjects were patients admitted to hospital with a first stroke and those managed as outpatients.

INTERVENTIONS

The effect on functional outcome after ischaemic or haemorrhagic stroke, tumours or infections, of correctly administered antithrombotic or other treatment; of time to scan and stroke severity on diagnosis by CT or MRI; on management, including length of stay, functional outcome, and quality-adjusted life years (QALYs), of the diagnostic information provided by CT scanning; the cost-effectiveness (cost versus QALYs) of different strategies for use of CT after acute stroke.

MAIN OUTCOME MEASURES

Death and functional outcome at long-term follow-up; accuracy of CT and MRI; cost of CT scanning by time of day and week; effect of CT diagnosis on change in health outcome, length of stay in hospital and QALYs; cost-effectiveness of various scanning strategies.

RESULTS

CT is very sensitive and specific for haemorrhage within the first 8 days of stroke only. Suboptimal scanning used in epidemiology studies suggests that the frequency of primary intracerebral haemorrhage (PICH) has been underestimated. Aspirin increases the risk of PICH. There were no reliable data on functional outcome or on the effect of antithrombotic treatment given long term after PICH. In 60% of patients with recurrent stroke after PICH, the cause is another PICH and mortality is high among PICH patients. A specific MR sequence (gradient echo) is required to identify prior PICH reliably. CT scanners were distributed unevenly in Scotland, 65% provided CT scanning within 48 hours of stroke, and 100% within 7 days for hospital-admitted patients, but access out of hours was very variable, and for outpatients was poor. The average cost of a CT brain scan for stroke was pounds 30.23 to pounds 89.56 in normal working hours and pounds 55.05 to pounds 173.46 out of hours. Average length of stay was greatest for severe strokes and those who survived in a dependent state. For a cohort of 1000 patients aged 70-74 years, the policy 'scan all strokes within 48 hours', cost pounds 10,279,728 and achieved 1982.3 QALYS. The most cost-effective strategy was 'scan all immediately' (pounds 9,993,676 and 1982.4 QALYS). The least cost-effective was to 'scan patients on anticoagulants, in a life-threatening condition immediately and the rest within 14 days'.

CONCLUSIONS

In general, strategies in which most patients were scanned immediately cost least and achieved the most QALYs, as the cost of providing CT (even out of hours) was less than the cost of inpatient care. Increasing independent survival by even a small proportion through early use of aspirin in the majority with ischaemic stroke, avoiding aspirin in those with haemorrhagic stroke, and appropriate early management of those who have not had a stroke, reduced costs and increased QALYs.

Assessing the resource implications of extending routine invitation to breast screening to women aged 65-67 years

UK breast screening policy currently restricts routine 3-yearly invitation to screening to 50-64 year olds. However, it is likely that routine invitation will be extended to 65-67 year olds in 2001. This paper first predicts the additional demand for breast screening as a result of this new policy by modelling the response to the 1998 invitation of women eligible for screening in 2001. The independent variables include (i) the woman's characteristics: her screening history; the deprivation score of the area she lives in; and (ii) the characteristics of the screening: whether the screening took place in a mobile van or at a static site; and time of the year. The modelling of attendance is quite successful in that most hypothesised variables have the expected sign. It is estimated that an additional 10829 women will be screened per annum. The additional invitation, screening and assessment costs are expected to be approximately pound350000 in 2001.

Symptom-based outcome measures for asthma: the use of discrete choice methods to assess patient preferences

This paper reports on an application of discrete choice modelling to the measurement of patient preferences over asthma symptoms. A sample of patients with moderate to severe asthma was asked to choose between a series of pairs of scenarios characterised by different combinations of asthma symptoms. Their responses were analysed using a random effects ordered probit model. The results implied that patients weighted some symptoms more highly than others. Discrete choice modelling proved to be a useful approach for developing preference based outcome measures, although the results show how, in contexts where preferences over health care outcomes based on symptoms or some measure of health status are involved, a conventional linear additive model may not always be suitable.

Estimating time preferences for health using discrete choice experiments

This study is the first to use discrete choice experiments to elicit inter-temporal preferences for health. Inter-temporal preferences with respect to one's own future health are compared with inter-temporal preferences with respect to others' future health. Discrete choice experiments are used to measure the relative importance of the duration of ill-health and how far in the future the ill-health occurs. Data were collected by postal questionnaire in the UK. The median implied rates of discount range from 0.055 to 0.091 for own health, depending on the period of delay, and from 0.078 to 0.147 for others' health. The implied discount rate varies with respect to age, self-rated health, and version of the questionnaire. The implied discount rates are broadly comparable with other published estimates using closed-ended methods. One concern is the large percentage of respondents with dominant preferences. This issue needs to be explored before adopting the approach of discrete choice experiments to elicit inter-temporal preferences.

Patient weighting of importance of asthma symptoms

BACKGROUND

Quality of life measures are increasingly important in evaluating outcomes in asthma. If some asthma symptoms are more troublesome to patients than others, this may affect their contribution to outcome measures. This study was designed to assess the relative importance of common symptoms in adults with asthma.

METHODS

A postal survey using conjoint analysis was performed in 272 adults attending hospital outpatient clinics with moderately severe asthma. Patients were asked to chose between "symptom scenarios" offering different combinations of levels of five common asthma symptoms over one week. Two versions of the questionnaire were used with identical scenarios presenting symptoms in different orders. Different patients answered the two versions. Regression analysis was used to calculate symptom weights for daytime cough, breathlessness, wheeze and chest tightness, and sleep disturbance.

RESULTS

Symptom order, percentage predicted peak expiratory flow (PEF), and symptoms in the week before the survey did not influence the choice of scenario. In both questionnaires patients were more likely to choose scenarios with low levels of cough and breathlessness than low sleep disturbance, wheeze or chest tightness. Regression weights for cough (-0.52) and breathlessness (-0.49) were twice those of wheeze (-0.25), chest tightness (-0.27), and sleep disturbance (-0.25). For 12% of patients cough dominated patient preferences, regardless of all other symptoms. Age was inversely related to weight given by patients to breathlessness.

CONCLUSIONS

The prominence of cough among other asthma symptoms was unexpected. Daytime cough and breathlessness had greater impact for patients than wheeze or sleep disturbance. Age influenced symptom burden, with younger patients giving greater weight to breathlessness than older patients. Conjoint analysis appears to be a useful method for establishing the relative importance of common symptoms.

The developing role of ecotoxicology in industrial ecology and natural capitalism

Images

The interface between molecular biology and cancer research

During the last thirty years, cancer research has been a remarkably fruitful resource for molecular biologists. Numerous fundamental discoveries in basic biology have come out of research into the properties of cancer cells; for example, the discovery of reverse transcriptase, RNA splicing and the protein kinases. Recently, information has started to flow in the other direction, and we are at last beginning to see molecular biology yielding discoveries of practical importance in the management and control of human cancer. Some of the past and possible future interactions of molecular biology and cancer research are discussed in this paper.

The efficient organization of blood donation: what determines the number of donors and donations?

This study investigates the influence of socio-economic characteristics on the number of donors in geographical areas in the North of Scotland as well as the influence of panel-specific characteristics on the number of donations obtained from each panel in the North of Scotland. Econometric models are applied to establish the relationship between these variables. The main findings were that there was no scope for identifying new areas in which to locate new panels or extend the activities of existing panels because the number of donors in an area was closely correlated with population. There was therefore little scope for recruitment of donors by focusing on any particular area. It was found that annual donations per panel are significantly influenced by the number of donors invited, the number of opportunities to donate and the average length of sessions. Combined with cost information, increasing the length of the session appears to be the most cost-effective means of collecting higher volumes.