Review of Radiotherapy Services Victoria 

Background

The Victorian Department of Human Services (DHS) has initiated this Review to provide the Government with an analysis of options for future development of radiotherapy services. ACIL Consulting Pty Ltd (ACIL) was commissioned by the DHS to undertake the Review. This report outlines options for the development of radiotherapy services in Victoria and makes recommendations based on a consideration of the social costs and benefits (which are broader than the clinical costs and benefits) of these options.

Following the terms of reference for the Review, the report addresses the features of demand for and supply of radiotherapy services, outlining the factors that influence these. This then underpins an analysis of costs, both operating and capital, and the role of the private sector consistent with the Government’s policy. Throughout these analyses, there is recurrent reference to, and examination of, issues concerned with cancer patient access to, and awareness of, treatment options which incorporate radiotherapy. These matters are central to the determination of appropriate strategies for further developing radiotherapy facilities in Victoria.

Against this background, options for the future development of radiotherapy are then presented, and analysed in terms of their costs and benefits, and trade-offs between competing objectives. The implications of these options have been explored to provide a framework for future decisions.

ACIL has examined the current availability and utilisation of radiation oncology services in Victoria, within the context of other oncology services and other elements of the acute health care system; these relationships prove critical to both the appropriate strategies and the assessment of the cost effectiveness of a significant increase in the utilisation of radiotherapy services. The Review has been based on extensive consultations with the range of stakeholders - health care professionals, government agencies and ‘consumers’ of the services - on submissions received through a public submission process and on detailed analysis of data drawn from a wide range of sources.

This Review builds on previous planning studies and reports in Victoria, notably the Lovell Inquiry in 1985, and the Cancer (Radiotherapy) Services Strategic Plan in 1990. More recently, the Australian Health Technology Advisory Committee (AHTAC) summarised the present knowledge of the key issues in the field of radiotherapy as they apply to Australia. Their report ("Beam & Isotope Radiotherapy", December 1996) contained several key recommendations to improve radiotherapy services in Australia, that have proved fundamental to this Review, notably that radiation oncology be organised through networks of services that provide coordinated, comprehensive care in multimodal settings, and that efforts are made to decentralise services to enhance access for consumers.

The Review was conducted in a four-stages - data collection, consultation, data analysis, and the development of recommendations and conclusions. The Review Advisory Committee, convened by the DHS to advise on in the conduct of the Review and to provide feedback on the work as it developed, proved an invaluable source of information and insight throughout the review. However, the conclusions and recommendations remain those of ACIL.

Statistical data and information have been obtained from key sources, such as the DHS, the Anti Cancer Council of Victoria (ACCV), the Commonwealth Health Insurance Commission (HIC), and service providers in Victoria. Consultation occurred with individual key informants, and through a public submission process. Submissions to the Review, and the results of analysis then formed the basis of a stakeholder workshop, in which key issues and options for future development were discussed in an open forum. Following the workshop, key conclusions and recommendations have been developed and are outlined in this report.

Working through the issues involved in the further development of radiotherapy services in Victoria requires addressing a range of complex trade-offs whose outcomes are heavily dependent on the value systems brought to bear. These include trade-offs between costs to the providers and costs to patients; between accessibility and use of clinically desirable treatments and the quality with which these treatments are delivered; between prescription and choice; between costs to the Victorian Government and costs to the Federal Government; between rapid deployment of probably more valuable treatment systems but with real risks needing to be managed and a slower, safer roll-out but possibly at the expense of some current patients and the overall costs of the health care system.

Under these circumstances there is no uniquely ‘best’ strategy. What is needed is a policy process which ensures progress towards outcomes which would be widely viewed as representing real improvement and which sensibly manages the risks involved; this requires that judgments be made and that those judgments be defensible - but not that they be beyond debate, nor the application different value systems. It also requires that the medium and longer term recommendations not be set in concrete because the process needs to be able to learn from the experience gained with implementing the initial recommendations. In effect, the recommendations, and the reasoning underpinning them, will need to be revisited over time.

Accordingly, ACIL has placed considerable emphasis on setting out the trail of information, analysis, judgment and logic which underpins the recommendations it has ultimately made. It is hoped that this will allow open assessment of the process used and the conclusions reached, that it will facilitate the implementation of a set of changes which will ensure improved outcomes over the next several years and that it will allow the flexibility to adapt to changing information.

This Overview is, deliberately, a fairly detailed document in its own right because of ACIL’s judgment that the review process, and the reasoning on which the conclusions are based, needs to be well understood as well as the recommendations if the subsequent policy process is to deliver the best outcomes. In fact, this Overview is more than just a summary of the conclusions developed in the report. It is deliberately structured to bring together the various conclusions and lines of argument - from a complex array of policy and data analyses - to present them as coherently as possible. This drawing together of the different lines argument in this way produces insights which are not explicitly highlighted in the main body of this report but which are as important to our final recommendations as the individual arguments.

Key Arguments and Conclusions

Based on information received and analyses undertaken during this Review, ACIL has developed the following arguments and broad conclusions:

• There is no basis for departing from the recent conclusion of the Australian Health Technology Advisory Committee (AHTAC, 1996) that, averaged across cancers, approximately 50 to 55 per cent of cancer patients would, from a clinical perspective, benefit from a treatment regimen which incorporates radiation oncology.
  
  
• This compares with current rates of utilisation in Victoria of the order of 42 per cent. In non-metropolitan areas, this rate is somewhat less, and estimated to be as low as 30 to 35%.
• It is noted that current clinical trends, and emerging US experience, suggest that the assessment of the clinically optimal rate of treatment will probably rise over the next decade, with up to 60 per cent of new cancers expected to be recommended for radiotherapy within 10 years.
• At the same time, both direct (though here far from definitive) and indirect evidence is presented which suggests that medical oncologists, surgeons and primary referrers currently believe that rates significantly below AHTAC levels are probably more appropriate - and this is seen as a significant impediment to increased utilisation.

• Access to radiation oncology services has several dimensions and there are grounds for concern with all of these:
  
  
• Two major factors which influence access are, firstly, the ease/cost with which these services can be obtained and, secondly, the extent to which both the referrers and the patients have sufficient, balanced information on which to base a decision as to an appropriate treatment regimen.
• Based on the information examined, it is almost certain that both factors contribute to the gap between the AHTAC guidelines and current levels of utilisation in Victoria. What is not so clear (and this has important policy implications) is the relative importance of the two factors.
  
  
• There are significant social and financial costs associated with the use of less accessible services and these are not, in general, fully compensated for by existing support arrangements.
• These costs can include some or all of lost time, direct travel and accommodation costs, child-care costs, pain and discomfort associated with travel, lost earnings (either of the patient or supporting family members) and removal from established family and community support.
• These costs can be functions of distance/travel times but also of time-of-day and day-of-week access to facilities which can especially influence costs such as child-care and lost earnings.
• Given the nature and duration of some radiotherapy treatments, especially non-palliative treatments, just the financial component of these costs could amount, in some cases, to several hundred, or even thousand, dollars, to which needs to be added the ‘value’ of the non-financial components.
• These costs need to be weighed against the costs of alternative treatments, such as more extensive surgery, but it is clear that there are many circumstances in which radiotherapy treatments would face a significant social cost penalty compared to some alternatives.
  
  
• These costs tend to be greater for individuals located at a distance from a centre suitable for delivering the necessary treatment, and in particular for individuals located in rural areas. This can be due to a combination of longer travel times, the likelihood of greater accommodation costs or less acceptable accommodation arrangements (proximity to family etc), and possibly less flexibility to arrange child care, to take leave of absence from work (especially from farms) etc.
• However, even within metropolitan Melbourne, both travel times and the wider opportunity costs of using these services can be large. Patients who live in the fast growing, outer suburbs, particularly in the Outer West and Outer South-East represent a substantial proportion of total demand and face travel costs and time costs which can be significant.
• The nature of these costs must also be considered relative to the costs of alternative treatments, because of the nature of the treatment regimen, including the common need for several treatments at a specialised facility.
• Given the larger population mass in Melbourne, it may well be that the aggregate access costs within Melbourne are comparable to those outside of Melbourne, even though the per patient costs would be lower within Melbourne.
• Access to real choice as to treatment options is likely to have social value in its own right, allowing individuals to better tailor treatment options to their needs and preferences and to experience a greater sense of control. Real choice is also dependent on both information and availability of services.

• There are, in many cases, significant social costs associated with the non-use of these treatments where their use is clinically indicated.
• Typically, these involve a significant quality of life component (such as disfigurement, pain and suffering, risk of recurrence etc) and a small component involving survival prospects - in most cases, the alternative treatments offer comparable reductions in risk of death due to the cancer, though some studies are suggesting more significant survival impacts.
• Even in the case of palliative care, some studies and anecdotal evidence indicate quite significant quality of life benefits in relation to cancers such as lung cancer.
• Quantification of these costs in ‘dollar’ terms is both controversial and constrained by available information. The confounding issue, alluded to earlier, between deterrence due to access costs and deterrence due to poor information makes it difficult to extract broadly based figures from historical patterns of choice, while the emotional sensitivity of the issues would work against valuation through direct survey methods.
• It is likely that patients who are deterred from the use of radiotherapy by the access costs attach a lower value to the perceived differential in expected treatment outcomes than do those who are not deterred - ie, any estimate of the average social cost differential between two treatment regimens would overstate the effective cost, due to the natural selection processes involved.
• Consistent and recurrent information provided in submissions, in the consultations and in the workshop, while largely anecdotal, did clearly reflect the view that these social cost differentials are, in some cases at least, far from negligible.
• Commonly cited were the cases of informed women, particularly in rural areas, choosing a mastectomy in preference to a lumpectomy and adjunct radiotherapy because of the perceived access costs.

• This said, access costs are real, and are for some time likely to remain higher for some radiation oncology treatments than for alternative treatment regimens with comparable survival profiles.
  
  
• In these circumstances, some ‘under-utilisation’ in the sense of average utilisation rates below clinically optimal standards is likely to be perfectly rational and appropriate for consumers;
• While equalising access costs as perceived by all potential users of the service, even if practical, would almost certainly not represent the best use of scarce health sector funds.
• ie, it is to be expected, and not interpreted as an indicator of system failure, that actual rates of utilisation will fall below current clinically optimal levels.
• A key issue is whether all of the access costs are necessary costs, or whether strategies can be implemented to reduce either the real or the perceived access costs to favour a more appropriate level of utilisation - possibly still below the AHTAC level.
• Clearly it is possible to redistribute some of the costs, particularly the financial costs; policies such as VIPTAS already have a significant redistributional effect.

• A number of highly relevant policy objectives for Victorian acute health services are identified in the 1996 Metropolitan Healthcare Services Plan, including:
• providing greater access to services closer to where people live;
• examining opportunities for private sector involvement; and
• developing models of service delivery which are cost-effective and maintain quality standards.

• Oncology services comprise cancer surgery, radiation oncology, medical oncology and haemoncology. Palliative care is also an essential component of care for many cancer patients. Many patients require combined modality treatment.
• There is wide agreement that each discipline should be part of a comprehensive cancer service, with close interaction between disciplines. In many areas of Australia, the lack of an integrated service has led to a situation where the modality used for a treatment for a particular type of cancer will often depend on the point of entry to the system.
• The care of the cancer patient requires access to a wide range of specialties. A comprehensive service requires the development of special groups with expertise in treating cancer in special anatomical sites such as breast, colon, neuro-endocrine, and head and neck.
• Full facilities for supportive care, tertiary pharmacy services, pain relief, rehabilitation and psychiatric assistance are also needed both within the hospital and community.

• Demographic and epidemiological trends indicate that there will be a need to treat significantly more cancer patients in coming years. This is just to sustain current levels of access and use.
• Indicatively, by 2006 the number of new cancers in Victoria will have risen approximately 22 per cent:
• Of this total of 22 per cent, most is due to the population growth and aging. About 8 per cent is due to the simple numerical increase in population growth, while 9 per cent is due to the increase in relative incidence due to the aging of the population. The remaining increase in incidence of about 5 per cent is due to several other individually minor factors, such as the impact of HIV/AIDS related cancers.
  
  
• More difficult to predict is the change in the proportion of cancers whose treatment includes radiotherapy.
• This proportion will be a creation of the policy environment as much as it is something to which the policy environment needs to respond.
• Any policy changes which reduce access costs or which alter consumer information or referral practice in a way which favours radiotherapy, can be expected to add to the demand for services.
  
  
• Given stated Victorian Government policy in relation to improving access, the implications of the conclusions and recommendations of the final AHTAC report, the evidence of the costs of underutilisation considered in this review (and the associated evidence that use of increased radiotherapy is unlikely to add to and may reduce whole of life treatment costs for cancers), ACIL has concluded that the policy environment both should and will contribute to a growth in utilisation rates.
• ACIL remains of the view that, taking into account unavoidable access costs, it is to be expected that actual utilisation rates will remain below then current assessments of clinically optimal utilisation rates;
• But ACIL recognises also that the assessed clinically optimal utilisation rate may well rise significantly over the next ten years.
• And has concluded that a year 2006 demand level in the AHTAC range of 50 to 55 per cent is reasonable and realistic for planning purposes.
• It will be a policy decision as to whether this demand is fully met or not (the earlier ‘chicken and egg’ argument aside).
• There may be real resource limits (professional/technical) on the rate at which utilisation can be expanded, though recommendations are made for addressing and monitoring these.
• However, assuming current AHTAC utilisation rates are to be met, this implies that the above increase in cancer cases of 22 per cent would translate into an increase in the number of radiotherapy treatments of between 45 and 60 per cent (combined growth in incidence plus increased uptake/utilisation), perhaps slightly reduced by a small trend towards fewer attendances per course of treatment.
• This increase in utilisation can also be interpreted as between 2,500 and 3,000 cancer patients each year receiving a clinically preferred cancer treatment involving radiotherapy rather than the alternative treatments which would currently be chosen.
  
  
• These demand projections can be broken down into estimates of the regional demand for local treatment, given an assumption regarding the extent to which there is a ‘leakage’ of demand out of the local region and towards a major metropolitan facility - due either to the referral of complex cases or the personal choice of patients.
• ACIL has assumed that leakage rate could be as high as 40 per cent, but this is likely to vary regionally and over time (with important policy implications discussed later) and is at present necessarily subject to considerable error. Analysis of leakage rates, and the consequent impact on the optimal distribution of capacity have been identified and discussed with reference to existing services.
• It is clear from views expressed to ACIL that a proportion of the population is likely to rate treatment at a major metropolitan facility as, in some sense, ‘better’ than that provided by a smaller local unit and would be prepared to face some additional costs to receive such treatment. This form of leakage may decline over time if the smaller units deliver a high quality service and this is documented and communicated through programs to monitor the quality of outcomes and through information provided to potential consumers and referrers.
• Furthermore, for some people in some parts of the defined regions, transport to inner Melbourne, or family support or accommodation in inner Melbourne, may be easier than with a regional centre, even one which is geographically closer.
• At the same time, the experience of several of the clinicians advising ACIL in this study would support a much lower leakage rate - of the order of 5 to 10 per cent. These assessments may be biased downwards to the extent that they are based on experience with those patients who first present at a regional centre - it is difficult for clinicians operating in peripheral centres to assess with accuracy the numbers of patients from their natural catchment who never present at the facility - but they certainly support the view that there is considerable conservatism built into the 40 per cent figure.
• Reflecting this range of views, ACIL has developed its analyses of machine demand based on a range of leakage rates between 10 per cent and 40 per cent.
• The lower figures certainly increase the demand estimates in the regions (and lower the estimates of demand in central Melbourne) but do not alter the overall picture sufficiently to have a major impact on the immediate recommendations.
• Estimates of machine requirements using a 40 per cent leakage rate are viewed as conservative, but amenable to adjustment as actual levels of leakage are observed. If observed leakage rates are much lower, and ACIL accepts that this is likely to prove the case, then this would underpin a more rapid roll-out over time, may justify the location of one or more machines in locations initially assessed as unlikely to justify the investment and would accelerate the relative decline in demand for treatment delivery in inner Melbourne.
  
  
• Further ‘medium scenario’ assumptions (discussed in more detail in the report) of a 50 per cent average utilisation rate and a capacity of 400 to 450 new patients per linear accelerator give an estimated total ‘demand’ of around 26 Linear Accelerators by 2006. Examination of the relative regional demand - expressed as the number linear accelerators - illustrates the catchment areas most likely to support a service through local demand:
• Of course, machines do not come in fractional sizes but there is some scope for varying machine throughput and the above figures, rounded down, give some indication (probably conservative using the 40 per cent leakage rate) of the number of units for which a viable level of demand might be achievable over the next ten years or so.
• Major centres, including metropolitan and Geelong may continue to service levels of demand above these levels by drawing from a wider area.

• Of these regions, Hume could expect demand up to 70% greater than indicated because of the adjacent NSW catchment (ie. 1.7 machines based on the 40 per cent leakage rate, approaching viability for a 2 machine unit - a quite plausible leakage rate of no more than 25 per cent would push the figure up to 2 machines).
• The assumed leakage rate has significant implications for the number of treatments being delivered by the major centres in inner Melbourne - and for the associated number of machines. Ignoring the lumpiness of the machines would suggest machine numbers ranging from 6.2 down to 1.7, depending on the level of referral. In practice, fewer machines are likely to be installed in the regions because of the problems of lumpiness, and this suggests demand for an additional 2 to 4 machines across these major centres.
• The above figures could vary considerably in future analysis - any strategies adopted in the short term which alter ease and cost of access will produce better information on which further analyses can be based - and should therefore be used as a guide, and not necessarily a constraint for future development.
  
  
• More generally, an increase in the number of treatments will require some mix of: new facilities; expansion of existing facilities; and increased utilisation of existing facilities.

• Increased utilisation of existing facilities through extended hours of operation appears quite feasible from a technical perspective.
• Extended operating hours are already common in some Victorian facilities, while facilities in other States, notably Queensland, offer models for further extension in a manner which could be expected to contain costs.
• From a financial perspective, reduced capital costs per patient need to be weighed against increased operating costs per patient. The latter are heavily dependent on the nature of the roster arrangements, on wage and salary agreements and on the extent of the ancillary services required.
• The cost relativities are heavily dependent on the assumptions made regarding the services which need to be available and the associated staffing. There is some scope for scheduling patients, depending on the severity of their condition and other risk factors, so as to reduce the staffing required during at least some of the extended hours operations
• Based on the available information, it seems unlikely that there need be any significant increase in provider cost per treatment (relative to expanding facilities) and there may be scope for reduction. Benchmarking undertaken by ACIL (based on approaches used in recent NSW modelling but adapted to incorporate our cost data) suggests scope for extending hours with little impact on provider cost per treatment - the base case model delivered a cost reduction of 3 per cent. Scheduling of patients to allow some reductions in the level of support services during the extended hours could deliver further cost reductions provided that throughput is achieved.
• The main factors influencing this analysis relate to the trade-off between savings made through more effective use of capital against cost premiums associated with labour costs. Most of the labour cost premium could be expected to affect radiation therapists, medical staff, and some administrative staff.
• Importantly, however, this expanded utilisation can be expected to lower patient access costs (for a minority of patients) if it delivers a better geographic spread of extended hours access - and, with this and the associated increase in utilisation rates, would probably also lower the costs associated with other forms of cancer therapy. Therefore, even if the extended hours involved a modest provider cost increase, this might still be cost effective.
  
  
• Differential rates of throughput, particularly the relatively higher throughput of private facilities in Victoria, point to some scope for increased utilisation of existing facilities even within standard hours.
  
  
• 3 machine private sector facilities in Victoria currently average in excess of 700 patients per year, in comparison with a public sector figure of around 500.
• However, care is needed in recognising that the teaching and research loads, and the more complex cases, handled through the public facilities necessarily entail some reduction in throughput and need not reflect lower efficiency. By international standards, throughput figures for both the public and the private sector in Victoria are high.