WORKFORCE CSIRO PUBLISHING

Australian Health Review, 2015, 39, 228–239 http://dx.doi.org/10.1071/AH14113

A systems life cycle approach to managing the radiology profession: an Australian perspective Seyedamir Tavakoli Taba1,4 BSc, MEngSc, PhD Candidate Simon Reay Atkinson1 BSc, MPhil, PhD, Associate Professor Sarah Lewis2 BAppSci (Hon), MEd, PhD, Associate Professor Kon Shing Kenneth Chung1 BCom, BSc (Hon), PhD, Lecturer Liaquat Hossain1,3 BSc, MSc, PhD, Professor 1

Complex Systems Research Group, School of Civil Engineering, Faculty of Engineering and IT, University of Sydney, Sydney, NSW 2006, Australia. Email: [email protected]; [email protected] 2 Medical Imaging Optimisation and Perception Group (MIOPeG), Faculty of Health Sciences, Brain Mind Research Institute, University of Sydney, Sydney, NSW 2141, Australia. Email: [email protected] 3 Division of Information and Technology Studies, RM113, Runme Shaw Building, Pokfulam Road, Faculty of Education, The University of Hong Kong, Hong Kong. Email: [email protected] 4 Corresponding author. Email: [email protected]

Abstract Objective. Although the medical system has expanded considerably over the past two decades in almost all countries, so too has the demand for health care. The radiology specialisation may be an early system indicator, being especially sensitive to changes in supply and demand in both rural and urban environments. The question is whether the new policies of increasing the number of radiologists can be a proper long-term solution for the imbalance of workforce supply and demand or not. Methods. Using system dynamics modelling, we present our integrated descriptive models for the supply and demand of Australian radiologists to find the actual gap. Followed by this, we pose a prescriptive model for the supply in order to lessen the identified imbalance between supply and demand. Our system dynamics models compare the demand and supply of Australian radiologists over 40 years between 2010 and 2050. Results. The descriptive model shows that even if the radiology training program grows at a higher rate than the medical training growth rate and its own historical growth, the system will never be able to meet demand. The prescriptive model also indicates that although changing some influential factors (e.g the intake rate) reduces the level of imbalance, the system will still stay unstable during the study period. Conclusion. We posit that Australia may need to design a new system of radiology provision to meet future demands for high-quality medical radiation services. We also suggest some strategies, such as greater development of radiographers’ role, are critical for enabling sustainable change over time. What is known about the topic? Long-term workforce planning for medical services at the national level has been very challenging for policy makers of the 21st century. The current demographic imbalance in the supply and demand of the Australian radiologist workforce makes it difficult to plan the effects of extra inflow of radiology students over time. What does this paper add? This paper discovers the current situation facing the Australian Radiology profession and identifies all the factors that influence the long-term matching of radiologist workforce supply and demand. This Australian case study adds to the current literature of medical workforce planning and its challenges. Moreover, this paper answers how the problem of workforce imbalance can be solved through a sustainable change over time. What are the implications for practitioners? Conventional responses in Australia and many developed countries in response to radiological services demand have been to increase the number of radiologists. However, our models and analyses show that this is not an all-inclusive long-term solution: merely increasing the number of radiologists will not result in a balance between supply and demand. Received 30 July 2014, accepted 6 October 2014, published online 17 December 2014

Journal compilation
AHHA 2015

www.publish.csiro.au/journals/ahr

A systems approach to managing workforce

Introduction There have been profound changes in global health care provision since the 1980s. This has seen the expansion of medical and health care provision across the population: increased levels of patient access, advances in medical technologies and diagnostic aids and increases in private and healthcare expenditure. At the same time, populations (at least in developed countries) have been aging and demand for health care and spending on health has started to outstrip economies.1 This has escalated earlier than forecast following the global financial crisis (GFC) and wide-scale austerity measures being applied to public and private social and healthcare budgets. The pressures may be found in existing healthcare budgets facing increased patient survivability, improved detection, aging populations and higher sociomedical expectations.1 Not surprisingly, the traditional funding model may be increasingly unsustainable, even more so for economies struggling to recover from the GFC. Australia entered the GFC stronger than many other countries and has been less affected. Nonetheless, examining the Radiology Workforce Report2 from a systems life cycle perspective of the Radiology profession, the potential denouement may have been deferred, not avoided. The current demographic imbalance in the radiologist workforce makes it difficult to plan the effects of the extra inflow of radiology students over time. In this regard, Health Workforce 2025 Report – Medical Specialties3 identifies ‘Diagnostic Radiology’ as one of the six medical specialities in Australia that ‘will be in shortest supply by 2025’, with a expressed workforce demand projected per annum growth at 3.8. A similar problem of workforce imbalance and workforce planning can be found in other medical services worldwide.4 As noted by Bloor et al., ‘The challenges of medical workforce planning are international, with similar policy imperatives, and similar neglect of practice variations and incentive structures’.5 In most cases, medical workforce planning is based on wrong assumptions, which aim to increase investments in the healthcare workforce by increasing its number.6 Overall, longterm workforce planning for medical services at the national scale is a key challenge for policy makers in our modern, everchanging society. Questions about how public health systems and communities can be better prepared to respond to public health demand have been receiving unprecedented public attention in the past two decades. The fields of medical care and public health are tightly interconnected: whereas medical care is concerned with individual patients, public health considers the ‘community as patient’, seeking to improve the health of the whole population.7 In this regard, developing alternative healthcare workforce educational and learning designs to addressing global challenges in public health is essential for future preparedness. The main objectives of this study were to investigate which factors influence the long-term matching of medical care workforce supply and demand and how the problem of workforce imbalance can be solved through a sustainable change over time. To answer these questions, we investigated medical care provision and the medical care workforce at the macro level, with a focus on the Australian Radiology system as the context of the study. Herein, we used the Royal Australian and New Zealand College of Radiologists (RANZCR) 2010 Workforce Report2 as one of our primary sources of demographic and modelling data,

Australian Health Review

229

acknowledging the relatively low response rate of participating radiologists (39.2%; n = 607). However, the RANZCR data have been acknowledged as a suitable source of data by Australian government organisations, such as Health Workforce Australia.8 This Australian case study adds to current literature of medical workforce planning and its challenges. Considering the relevant literature at national and international levels, we examine factors like the inflow of students, the part-time choices of the workforce, the [current] aging profiles of the workforce, demographic profiles of the potential patients, societal factors including economic and environmental disturbances, emergence of subspecialisations in our study and modellings. In this paper, system analysis was undertaken to situate the Radiology profession in its wider medical and economic setting. In doing so, we examined the system as a whole before attempting a more detailed identification and classification of a potential model. This approach is the reverse of many performance-based models seeking to standardise as a means of optimising at the operational and unit level. System-level identification allows for classification and the development of appropriate strategies for improving productivity.9,10 Standardisation alone is generally used for optimisation; for example, by vertical integration and competitive resource constraints between supply chains.11 This paper presents a new cross-disciplinary research that looks at the healthcare system and applies a system dynamics method to healthcare workforce provision as a key global challenge. In the context of this paper, the term ‘medical imaging services’ refers only to those services that are categorised as Diagnostic Imaging Services under the Australian Medicare Benefits Schedule (MBS).12 The rest of the paper is organised as follows. We first discuss the background to the work and identify the current situation facing the radiology profession. We then describe our methods of using system dynamics modelling. We review where this approach has been applied previously and explain why it is applicable to our case. We present our integrated models for the supply and demand of the Australian Radiology workforce, followed by system analysis and our interpretations. Finally, we demonstrate our prescriptive model base on numerical analyses and suggest some strategies for enabling sustainable change over time. Background: a radiologist macro systems model The Radiology specialisation has been identified as being an early indicator (a canary) for the health of the medical profession.13,14 In Australia, to become a radiologist one must have a graduate degree in medicine. Newly qualified doctors then need to complete 2 years of general medical work in a hospital. After this period, these junior doctors are eligible to apply for the highly competitive 5-year radiology training program, accredited by RANZCR. Representing approximately 6% of graduating medical students in Australia, the 12-year education program (from commencing medical studies) is one of the longest specialist professional pipelines.2 Compared with a graduate employee, radiologists spend three times as long in higher education (Table 1) and, for every year in education, 1.4 years working after graduation. The radiologist workforce is relatively expensive and difficult to plan because radiology education takes

230

Australian Health Review

S. T. Taba et al.

Table 1. Education, higher education and working life parameters

500

Education Higher Working Education : (years) education life working (years) (years) life

450

2000

17 20 25

4 7 12

43 40 35

40% 50% 71%

9% 18% 34%

25 years for a relatively short working life of 35 years. The radiologist workforce is more difficult to plan for, because the longer educational period for radiologists makes predictions of the outflow from the educational system less reliable. Given these constraints, the profession may be highly susceptible to minor changes in recruiting and retention flows. It is also possible that a 6% radiologist extraction rate will represent a long-term constant, in which case, based on US figures,15 for every extra radiologist an additional 20 medical students (allowing for drop-outs) would need to start at medical school. However, such a simple measure may not, in reality, reflect those medical students actually pursuing a career in radiology. A question the Radiology profession may ask is, ‘What is the face of the Radiology profession today and what does the profession think it should be?’ Noting the sensitivity of the profession to changes in its supply pipeline, another question may be, ‘What is a sustainable professional age profile?’ From the RANZCR Radiology Workforce Report,2 it was possible to model the radiologist age profiles for 2000, when the 35–44 year age group was the largest, compared with 45–54 years in 2010, when the average age was 49.6 years (median 48 years). In 2000, there were 1148 radiologists compared with 1543 in 2010. Fitting this profile into a 35-year working life cycle from 30 years (the earliest age of entry on qualification) to 65 years (on retirement), the two profiles indicate not only an aging of the population, but a possible reduction of opportunities for newly qualified radiologists entering the profession. In 2000, almost 90 positions were available for newly qualified radiologists in the 30–34 years age bracket, with this figure reducing to approximately 80 in 2010. This is consistent with the RANZCR Report,2 which indicated that, despite increasing demand, opportunities for newly qualified radiologists were reducing. A combination of factors came together that, although meeting overall demand, constrained opportunity for newly qualified radiologists. Specifically, the late Baby Boomers and early Generation X who had entered the profession in the 1990s (as the profession expanded) were aging within the profession. At the same time, part-time positions were increasing, meaning that although there was increased flexibility of employment, there were also reducing job opportunities as one or more radiologists held onto posts that may otherwise have become available. Figure 1 shows how the profiles changed between 2000 and 2010. More detailed analyses showed that an Australian radiologist over a full working career of 35 years may have between five and 10 jobs. Thus, on average, a radiologist may practice for a maximum of 7 years at each job position and/or role, but not necessarily change their geographical job location. We suggest that the growth in the profession in Australia has probably been sustained by recruiting into the profession later (after 30 years

2010

400

Number of radiologists

Graduate Medical Radiologist

Higher education : working life

350 300 250 200 150 100 50 0 30–34

35–39

40–44

45–49

50–54

55–59

60–64

Fig. 1. Radiologist age/face profiles 2000 and 2010.

of age) and from abroad, which may be unsustainable in the longer term, noting that, in Australia, the average age of a first year medical student is 24 years (after they have completed a 3–4-year undergraduate degree). RANZCR suggest ambitious targets based on various different projections for the profession and Australian population. Based on its six different studies, 2 we consider increasing radiologists numbers from 1543 to 2062 by 2021. A question then arose as to how this may be achieved based on the need to simultaneously manage both growth and decline (due to retirement). Figure 2 considers three faces: (1) the 2010 profile (average age 50 years); (2) the 2021 population, allowing for the aging of the Baby Boomer (born 1946–1961) and Generation X (born 1962–1974) populations with the same numbers as 2010 (Scenario 1) and, third, allowing for population growth to 2062 based on the same profile (Scenario 2). In both instances, the 2021 profiles show the radiologist face continuing to ‘grey’, from 50 to 53 years. The impact of an ‘aging tail’ on opportunities at the start of career is significant, which may act unintentionally as position or job blocking in later years. Positions reduce from approximately 90 in 2000 to 80 in 2010 and to 70 in 2021. Only by growing the population to 2062 in 2021 do start-of-career opportunities increase (to 100). A complex system necessarily manages both growth and decline and ‘hunts’ for its equilibrium positions.16 Nevertheless, it cannot always grow in order to sustain vacancies and opportunities at the start of career. This indicates that the current population may be inherently unstable and potentially unsustainable (either interstitially or existentially) over the longer term. Two questions we considered were: (1) what could a sustainable face for the radiologist profession look like; and (2) how can this be achieved? It is suggested that a sustainable face may be 45 years. ‘Shaping’ the profession accordingly indicates it would be necessary to carefully manage both start- and end-of-career opportunities and/or positions (Fig. 3). To grow and re-profile over the 10 years to 2021 would require a targeted repositioning

A systems approach to managing workforce

Australian Health Review

600

231

700 2010

500

600

2021

500

Turnover

Number of radiologists

2021

400

300

400 300 200

Joining and leaving Leaving

100 200

Joining

0 2011 100

2013

2015

2017

2019

2021

Year Fig. 4. Radiologist recruiting/voluntary professional retirement program to manage growth and decline to 2021.

0 30–34

35–39

40–44

45–49

50–54

55–59

60–64

this time, this would require simultaneously managing an increase in recruiting and voluntary professional retirement (Fig. 4). This will be politically highly sensitive and made harder by constraints at federal, professional and state level, including changing part-time and full-time positions, gender profiles and the balance between the private and public sectors. Superimposed upon these changing factors is the need to preserve expert competency levels17 within the profession.

Age (year) Fig. 2. Radiologist age/face profiles 2010 and 2021.

600

2010 2021 (2010)

500

Number of radiologists

2021 (Pr)

400

300

200

100

0 30–34

35–39

40–44

45–49

50–54

55–59

60–64

Age (year) Fig. 3. Radiologist new age/face profiles 2010 and 2021.

and/or redundancy program paring the current 50–54 years cohort at a higher rate (approximately threefold) than historical voluntary professional retirement would suggest and at a rate 1.4-fold the historical rate for the current 45–49 years cohort. At the same time, it would be necessary to increase start-of-career opportunities and those in the 35–39 and 40–44 years cohorts by 50% and in the 45–49 years cohort by 25%. Such a program is not without precedence, but may not be addressed at anything other than a systems level. With a turnover of radiologists entering and leaving the profession of approximately 30% a year (necessary to sustain an average professional life of 7 years), managing the status quo will be a challenge. In addition, if the population is to grow over

Methods System classification Researches have considered the imbalance between supply and demand in the Australian medical system,18–20 and specifically in Radiology,21–23 but not necessarily at the system level. Two reports from RANZCR in 2000 and 2010 anticipated a potential future crisis emerging in the radiology workforce.2,24 The 2010 report discussed the demands population growth, population aging and imaging services would pose in addition to supply factors, such as medical graduates, working hour reductions and the early retirement of Baby Boomers.2 To meet demand, Australia is planning for an expansion of medical schools and Radiology programs.2 System dynamics modelling Forrester25 was the first to structure a method for modelling and analysing the behaviour of complex social systems that later became known as system dynamics. He originally founded his new approach on ‘computing technology, computer simulation, strategic decision-making and the role of feedback in complex systems’.26 In 1968, he redefined the basis of this approach as a closed boundary around the system, feedback loops (basic structural elements within the boundary), stock (level) variables representing accumulations and flow (rate) variables representing activity.27 System dynamics models are usually demonstrated by influence diagrams, where arrows represent influences between the different factors. System dynamics methods have been applied to many social, economic and environmental systems,28–31 the examination of which requires a holistic view and an understanding of

232

Australian Health Review

interrelationships via feedback loops.32 The method has been very well applied to the context of social care and healthcare33,34 and, more specifically, to the supply and demand of these systems35,36 in recent years. In the rest of this section, we explain how we developed our system dynamics models by examining all the variables influencing the Australian radiology system as a whole. Demand on radiologists Population growth in Australia is dependent on its politically stable economy and has been considerably higher than natural increases in the past 5 years. The Australian Bureau of Statistics (ABS) considers three different scenarios (A, B and C) for population growth based on estimates of fertility, life expectancy at birth, net overseas migration and net interstate migration. Scenario B (1.2%) is based current trends, whereas Scenarios A (0.8%) and C (1.6%) reflect wider bounded estimates. Scenario B is lower than the historical average,2 and the historical average of 1.4% population growth is used in this model as being more realistic. The ABS has noted that ‘in 2011, the early Baby Boomers turned 65’ and ‘projected that those aged 65 years and over would account for 14% of Australia’s population in 2011 and would increase to 20% of the population in 2030’. 37 The Economic Implications of an Ageing Australia report38 anticipated that the population aged over 65 years is likely to triple during the period 2005–45, with the population aged >85 year increasing fourfold during the same period. This model considers that the expected growth in the number of older people in the Australian population will result in 1% more demands on radiology services each year. The number of radiologists per million population varies widely in developed countries. In 2004 the number of radiologists per million population was 61 in Germany, 64 in Australia, 95 in the US, 125 in France, 156 in Italy, 194 in Denmark and 226 in Greece.23 The current ratio of 70 radiologists per million population in Australia is below the average of 100 in other developed countries.2 A realistic target for Australia may be 100 radiologists per million population. Although developments in new treatments, prognoses and care have decreased the burden of some diseases, population aging and environmental and lifestyle factors have increased the burden of others. Lifestyle factors, such as less physical activity and increasing obesity,39 and environmental factors are changing life cycle profiles and life expectancies. Some studies in Australia show that the rates of chronic disease, prostate cancer, diabetes, anxiety and depression are increasing.20 These figures point towards a growing demand for diagnostic and interventional health care. The pattern of usage of medical imaging services is changing, with an increase in defensive medicine techniques (with the desire to reduce the risk of liability), as well as a rise in medical litigation.40 We consider in the future that technologies and procedures will provide improved levels of treatment for many diseases, in which medical imaging (and particularly radiology) will play a key role. Conversely, demand for medical imaging services is also driven by factors such as doctors’ preferences, patient expectations and budgets. Although the population growth in the Australian population over the period 2000–09 was approximately 15%,2 Medicare data

S. T. Taba et al.

showed a 51% increase in the demand for medical imaging services over the same period; this means that medical imaging has experienced a 36% net growth in 9 years or a 3.5% net growth per annum. In this model, we assume that medical imaging services will have at least a 1% net growth per annum for the next 40 years. We know that the 3.5% growth in medical imaging services includes population aging and burden of disease impacts, as well as services growth for other reasons. Yet the total 2.2% growth in workforce demand based on these three influences in the model is much less than 3.5% growth in the past and this adds validity to our estimate of demand in the model (Fig. 5). Supply of radiologists Inflows to the healthcare workforce system are from national training programs and immigration. In response to health workforce shortages, Australia recently planned for significant increases in medical school graduates. The health workforce enlarged by 11.6% between 1996 and 2001, and by 22.8% between 2001 and 2006.19 In 2009, there were 1914 Australian graduates from national medical schools and 478 international graduates; in 2010, there were 2264 national and 512 international graduates.2 Our model considers medical graduation by positing future growth based on historical growth. Currently, approximately 6% of medical graduates commence Radiology training as a speciality. Nevertheless, the radiology training program would still need to grow if it is to accommodate a fixed input of 6%. ‘The ability of both the training program and the industry to absorb such numbers is questionable, considering that the current intake of first year registrars is ~68–70’.2 At current rates, the system (training program and the industry) is able to absorb approximately 60% of first year registrars. By increasing the number of medical graduates, this proportion may, in fact, decrease. The model assumes that there is the same growth rate for Radiology speciality training as for medical training, so the uptake rate stays constant at 60%. Gender proportions and the workforce Feminisation of the medical workforce appears to be a global trend. Research over the past decade41–45 has examined trends towards feminisation in medicine. In this regard, we consider, after Douglas,46 Ferguson47 and Fondas,48 ‘feminisation’ to be: . . .the spread of socio traits or qualities that are traditionally associated with females to things (e.g. technologies) or people (professions) not usually described that way; including the shift in gender roles towards a focus upon the feminine, as opposed to the pre-modern cultural focus on masculinity. Studies in the UK, US, Canada and Australia in specialist areas show that the proportion of women in medical schools has increased gradually over recent decades.43,49,50 Today, 60%–70% of students entering medical schools in Europe are women.51 A study published in 2001 indicated that the gender ratio in Australian medical schools was then 50 : 50.52 The gender ratio of radiologists in the US has been studied for many years. A 1995 study showed that the proportion of female radiologists was inversely correlated with age: 23% for those 44 years of age. Overall, only 13% of professionally active radiologists in the US in 1995 were women.53 Later, Lewis et al.54 analysed data from the American College of Radiology’s 2003 survey and stated that ‘in the USA, women represented 24% of all radiologists in training (residents and fellows) and 18% of post training, active radiologists’. Owen et al.53 also identified that working part-time was more popular with women (16%) than men (7%). That study suggested that women start working part-time sooner than men, at 39 and 63 years of age, respectively. In Australia, the RANZCR 2010 Radiology Workforce Report found that women accounted for 17.1% of radiologists in 1998 and 23.6% in 2010, with an anticipated 34% of radiologists in 2020 being women. Based on historical data, we defined a function in the model (Fig. 6) to consider the future growth of female radiologists reaching a maximum of 65% female by 2050 based on current European trends. Feminisation of the Radiology population may have as a significant impact on the system as it does on the face and age of the population. On average, the working life of a female doctor may be 60% that of male doctors.18 From the RANZCR report,2 17.3% of Australian female radiologists may be working parttime, compared with 6.7% of male radiologists. In Canada, Weizblit et al.45 argued that female doctors work on average 47.5 h per week, compared with 53.8 h per week for male doctors. They also somewhat controversially concluded that, ‘if the gender-specific work patterns described in the present study persist, an overall decrease in doctor productivity is to be anticipated’.45 Similarly, McKinstry55 talks about a ‘Time Bomb in the near future when older mainly full time (male) doctors retire’ and that a shortfall will occur in specialities in which the majority is part-time (female) doctors. Gould51 coined the phrase ‘leaky pipeline’ to label a situation in which large numbers of women Basic population 22 342 000

start out on a career, only for many to give up after training and few ever making it to senior positions. The primary reasons given are the different work patterns of male and female doctors, such as part-time versus full-time work, working out of hours, parenting, retirement age, the development of the profession55 and unequal distribution in specialties seen to be ‘family friendly’, such as primary care rather than surgery.56 The trend towards reduced working hours is increasing among Australian radiologists. Studies show that, in 2010, 34% (20% in 2000) planned to decrease their working hours, whereas only 8.7% (16% in 2000) planned to increase them.2 Brooks et al.18 argued that, beyond the feminisation effects on the medical workforce, male doctors also prefer to work less than before. Assuming that a part-time radiologist is working half of a fulltime one, the full-time working hours are calculable at 47.17 and 43.5 h per week in 2000 and 2010, respectively. The function in the model expects working hours in the future to be based on a dynamic full-time working profile (but takes up 37 h per week as a minimum full-time equivalent). Population growth and aging is likely to place potentially extraordinary demands on health services. Changes to the workforce may also have the same demographic impact. The ABS estimates that ‘Australia’s civilian labour force aged 15 and over to grow to 10.8 million in 2016, an increase of 1.5 million or 16% from the 1998 labour force of 9.3 million’. 57 Yet, the average annual growth rate of 0.8% between 1998 and 2016 is less than half that for the period 1979–98. Aging suggests that a smaller proportion of the population will participate in the workforce in the future than in the past. Like many other occupations, it is broadly foreseen that retirement among aging doctors will lead to a clinician undersupply in the future.58 As the Baby Boomers leave the workforce in the 2020s, and the retirement rate potentially increases, the model considers

Population

Radiologists per million

39 113 371.678

100

Population aging impact

Polluted environment

0.01

0.001

Population growth rate

Demand based on population

0.014

3911.337

Life style 0.002

Population growth 547 587.203

Defensive medicine 0.005

Payment mechanisms

Work force demand

Burden of disease impact

4032.589

0.003

0.002

New modalities 0.005

Work force hour demand 181 466.488

Imaging service growth 0.018

Standard work hour

Patient expectations 0.002

Doctor preferences

233

45

0.002

Fig. 5. Demand for radiologists and its related working hours in 2050.

234

Australian Health Review

S. T. Taba et al.

No longer registered

Radiologists aging

4236.287

Emigration Departure from work for other reasons 0.005

0.005

Normal Retirement

Outflow

Radiologists aging impact

128.423

0.01

Baby Boomers Retirement

Retirement impact 0.038

Basic radiologists 1543

Mortality

Average retirement age

0.002

65.46

Intake

Active work force

136.296

New trained radiologists

2558.188

3500.983

Active work force hour 88 211 266

Working hours 34.482

Radiology graduates

0.6

227.16

Medical graduates 3786

F Medical training growth

Life style change

Female ratio

F Medical training growth history

37

0.652

F Female radiologists growth F

Intake rate

Female radiologists growth history

F Life style work hour F Life style work hour history

Choosing radiology rate 0.06

Fig. 6. Supply of radiologists and its related working hours in 2050.

200 000 180 000

Workforce hour

1% growth in the retirement rate after 2020. It also assumes that the current average intended retirement age (65.46 years) stays constant in the future. The annual loss from the radiologist workforce for other reasons is difficult to estimate, but the outflows that the model considers assumes emigration (0.5%), mortality (0.2%) and departure rates from the workforce for other reasons, such as taking up an administration role (0.5%). Results and Discussion

Workforce hour demand

140 000 120 000 100 000 80 000 60 000 2010

2015

2020

2025

2030

2035

2040

2045

2050

Year

Fig. 7. Radiologists workforce demand and supply (2010–2050).

200 000 180 000

Workforce hour

Figure 7 shows the demand and supply of radiologists in a dynamic model over 40 years between 2010 and 2050. It appears that even if the radiology training program grows at a higher rate than the medical training growth rate and its historical growth, the system will never be able to meet demand. This suggests that the 6% intake rate of first year registrars will be insufficient and industry absorption will need to grow at a higher rate than the Radiology specialist training program. If the system is not able to increase the intake rate faster than the new radiologist supply rate (coming from national training programs and immigration), the profession may face two crises. First, supply and demand at current rates will never balance (Fig. 7); second, the number of trained, qualified radiologists who fail to enter the profession through the supply pipeline (unemployment in the occupation) may account for more than 3500 specialists over 40 years (Fig. 6). Figure 8 shows that the difference between demand and supply in active working hours is more critical than simply increasing the number of radiologists. The effects of factors such as feminisation and lifestyle changes are increasingly clear. We suggest that a side effect (more pressure on active radiologists) of this imbalance may result in the reduction of the quality of imaging services to patients and to profession employers, a potential ‘third crisis’.

Active workforce hour

160 000

160 000

Active workforce hour Workforce hour demand

140 000 120 000 100 000 80 000 60 000 2010

2015

2020

2025

2030

2035

2040

2045

2050

Year

Fig. 8. Radiologists working hours demand and supply (2010–2050).

The US faces the same supply and demand issues.59,60 Although the number of US radiologists increased by 1%–1.5% annually between 1992 and 2002, the workload also increased by 26% (~2.33% per annum).61 More dramatically, another study

4200 4000 3800 3600 3400 3200 3000 2800 2600 2400 2200 2000 1800 1600 1400 2010

235

Workforce demand Active workforce

2015

2020

2025

2030

2035

2040

2045

2050

Year

Fig. 9. Prescriptive model to change supply (2010–2050).

200 000 180 000 Active workforce hour

Workforce hour

in 2001 stated that the number of radiologists entering the profession was growing by 2% per annum, whereas demand was growing by 6%.22 Conversely, in a survey of the healthcare workforce considering undersupply in the US,62 the vacancy rate (unfilled positions) for imaging and/or radiology technicians was the highest for all allied health professions. We believe Australia is facing a similar situation; however, the solution may be different from that in the US because the healthcare systems differ significantly. The number of Australian healthcare professionals increased by 14% between 2001 and 2006 compared with a total workforce growth of 10% for the same period.63 However, this may not be enough to balance largescale demographic changes and anticipated demands and attractors or inducements to joining the profession. The training program and the industry absorb just 0.6 of first year registrars at current rates. Thus, the descriptive model (Fig. 6) used the take up rate of 60% for numerical analysis. Later, we developed a new prescriptive model to determine how changing the intake rate affects supply. The new model assumes that, every 5 years, the intake rate has increases as follows: it rises from 60% in 2010 to 70% in 2015, to 80% in 2015, to 90% in 2020 and to 100% in 2025 and later. Figure 9 shows the results for such an analysis, where the gap between supply and demand declines gradually. The effects are not perfect and the model attempts to be realistic rather than idealistic. In this regard, we also performed similar analysis for the active workforce hour. In addition to increasing the intake rate, the new model undertakes the following assumptions. First, the trend of working hours among radiologists will not decrease and stays constant at current figure of 43.5 h per week. Second, the feminisation of the workforce will not follow the European trends and grows gradually to a maximum of 50% by 2050. We mentioned in previous sections that the gender ratio of Australian medical students has already reached 50 : 50 and the upcoming generation of doctors is changing the face of profession. Figure 10 shows that supply (of working hours) will not meet the demand even if we apply the prescriptive model. A conclusion is that the traditional models of health care, specifically in radiology, are unsustainable based on growing demand for healthcare services. Australia may need to establish a new model of radiology service provision to meet future demands for high-quality medical radiation services. Because of unique interstitial demands on the Australian healthcare system, this model needs to consider the development of patient care, patient safety and quality of service in addition to increasing the number of radiologists. Therefore, a new model of radiology services may need to address change based on patient focus, skill sets and competencies. Conventional responses in Australia and many developed countries in response to radiological services demand have been to increase the number of radiologists. However, a system-level solution may also consider the ‘development of the radiographers’ role’64 in the healthcare system. There is a tight coupling between radiologists and radiographers. In their study on pulmonary nodule detection from chest radiographs, Manning et al.65 showed that both professionals are able to get comparable scores. In another study, selected trained magnetic resonance radiographers and consultant radiologists reported on

Australian Health Review

Number of radiologists

A systems approach to managing workforce

160 000

Workforce hour demand

140 000 120 000 100 000 80 000 60 000 2010

2015

2020

2025

2030

2035

2040

2045

2050

Year

Fig. 10. Working hours demand and supply for the prescriptive model (2010–2050).

specific magnetic resonance imaging examinations of the knee and lumbar spine at a similar level.66 A comprehensive study of the literature by Brealey et al.67 provides an evidence base showing that trained radiographers can report plain radiographs as accurately as radiologists in clinical practice. In many other countries, radiographers have a broader role in the healthcare system than in Australia; for example, in the UK radiographers have been one of the two independent BreastScreen readers since the early 1990s and have reporting rights through the National Health Service.68 However, reporting medical images by radiographers is not current practice in Australia, although there is robust debate about the evolution of advanced practice pathways for radiographers by both radiographers themselves and radiologists, who generally, as a professional group, are opposed to this role expansion.69,70 In this regard, and adding further to polarised opinions, Health Workforce 20253 identifies ‘extended scope of practice of radiographers’ as a necessary reform to ensure ‘a sustainable workforce’ for the future. The Australian health workforce system has had a shortage of radiographers over the previous decade running at similar levels to those for radiologists. Australia increased the number of new radiographers by 42.6% between 2001 and 2005.71 After 2007, medical radiation science (MRS) programs expanded and offered more places at universities. Although students with high entrance scores enter the Australian undergraduate radiology programs, it appears that significant numbers retire or leave the profession soon after graduation, or use their undergraduate qualifications for postgraduate study. This leakage may be due,

236

Australian Health Review

in part, to the lack of clinically oriented career development opportunities.72 It has been suggested that investing in development has the potential to improve radiographers’ job satisfaction and staff retention, which may be as important as recruitment. Studies also show factors such as professional opportunity, recognition of accomplishments and variety of work affect recruitment and the retention of health professionals more than existential factors like low pay and working conditions.13,14 The other reason for leakage may be the oversupply of radiographers. The idea of engaging radiographers in roles traditionally occupied by radiologists is not new. Indeed, radiographers informally have carried out elements of image interpretation and examination reporting in many developing and developed countries when radiologists were not available. More than 40 years ago, Swineburne73 wrote about ‘the possibility of improving diagnostic X-ray services and alleviating radiological workloads by using radiographers or other non-radiological staff to distinguish normal from abnormal films’. We emphasise the successful implementation of this concept, namely authorising diagnostic radiographers to take new practice roles beyond traditional ones. We also suggest caution. At the unit and operational level this may offer potentially attractive opportunities for optimisation and fractionation as a way of reducing costs by reducing skill contribution and, thereby, investment in the individual.74 At the system level, this type of ‘just-in-time’75 approach can cause problems to the development of skill sets over time, particularly in a tightly coupled system such as exists between radiographers and radiologists. That is, at the accountancy–consultancy level, radiographers appear to be an increasingly attractive (and cheaply) available option, but at the actuarial level the development of expert radiologists becomes increasingly constrained. This essentially leads to resource constraints,76 vertical integration77 and competition between suppliers (radiologists and radiographers),78 rather than for best provision of care to patient and healthcare employer and, given the backdrop of reducing budgets, to hypercompetition.11 Ultimately, this may create an unhealthy situation where the trust necessary for collaboration79 is broken and the profession and system as a whole is less able to problem solve in the interests of the patient and employer.80,81 That is, these types of performance-managed approaches (i.e. health care by targets) have often, in practice, led to reductions in health care provision. The appropriate application of part-time workers, if managed at the system level, could enable the necessary flows into and out of the profession while maintaining on-entry positions and the generation of expertise and experts later in the profession. Fragmented and fractionated82 as the management of the profession may have been to date, the projected increase in parttime workers is likely to negatively impact these positions through job blocking. Conversely, the number of cases read per year by radiologists may be a more reliable metric of current expertise (higher observer performance).83 There are different minimum volumes of cases required to be read to stay current in the profession (e.g. in mammography, 5000 mammograms need to be read per year in UK, compared with 2000 in Australia and 500 in the US).84

S. T. Taba et al.

System adaptation may improve system performance as a whole by allowing radiologists to concentrate more on diagnostic processes and allowing for collaboration between radiographers and radiologists. This collaboration may also lessen the rate of errors (both false positive and false negative) in cases. In this way, professional development, including taking on new clinical roles, may improve workforce retention. Significant differences in the number of radiologists per million population in different countries may also be impacted by current application and employment regimes. In Germany, approximately 70% of imaging studies are done by non-radiologists,85 and this may be a model that a systems-level approach may recommend. However, internationally there is great debate and confusion about the actual workforce attributed to ‘acquisition’ and ‘reporting and/or interpretation’ of medical images, with the nomenclature being often intertwined and misused.86 In Australia, there are no available statistics on image interpretation by radiographers because this is not a currently authorised pathway leading to a rebate through the MBS. The Medical Radiation Practice Board of Australia does not list formal written reporting of medical images leading to diagnosis as a professional capability of diagnostic radiographers.86 To increase system performance in various clinical departments, identification and classification of patient characteristics may also be necessary, specifically in ‘redesigning [the] scheduling scheme’.87–89 After Walter90 and Huang and Marcak,91 we believe patient classification in a hospital radiology department may help improve patient access to care and to optimise the use of medical resources, including balancing the time required for imaging by both the radiology workforce and patient. Huang and Marcak91 suggest that reclassification of patient groups by a new proposed scheduling scheme (using a decision tree technique and algorithm) ‘reduces patient wait time as much as 71%, increases the radiographer utilisation by at most 83%, reduces overall cost by as far as 54%, and improves patient access by 1.25 times than current schedule capability’. Conclusion This paper presents new cross-disciplinary research that applies a system dynamics method to healthcare workforce provision. It identifies the radiologist specialisation as an early indicator: sensitive to changes to supply and demand. By assessing the current situation facing this profession, we present our descriptive and prescriptive models and related numerical analyses for the supply and demand of Australian radiologists. In all our analyses, achieving the demand of radiology is very difficult, if not impossible, using conventional responses of just increasing the number of radiologists. From our point of view, traditional models of health care, specifically in radiology, are unsustainable based on growing demand for services. By identifying the tight coupling between radiologists and radiographers, we posit, after Smith and Baird,64 the development of the radiographers’ role in the health system as a potential solution to make the whole system sustainable. We also suggest that appropriate application of part-time workers and reclassification of patient groups are necessary to achieve this goal.

A systems approach to managing workforce

Competing interests None declared. References 1

2

3

4 5 6

7 8

9

10

11 12

13

14

15

16 17 18

19

Getzen TE. Health care is an individual necessity and a national luxury: applying multilevel decision models to the analysis of health care expenditures. J Health Econ 2000; 19: 259–70. doi:10.1016/S01676296(99)00032-6 The Royal Australian and New Zealand College of Radiologists (RANZCR). 2010 RANZCR radiology workforce report. Canberra: RANZCR; 2011. Health Workforce Australia. Health workforce 2025, volume 3, medical specialties; 2012 Available at: http://www.hwa.gov.au/sites/uploads/ HW2025_V3_FinalReport20121109.pdf [verified November 2014]. Zurn P, Dal Poz MR, Stilwell B, Adams O. Imbalance in the health workforce. Hum Resour Health 2004; 2: 13. doi:10.1186/1478-4491-2-13 Bloor K, Hendry V, Maynard A. Do we need more doctors? J R Soc Med 2006; 99: 281–7. doi:10.1258/jrsm.99.6.281 Maynard A. Medical workforce planning: some forecasting challenges. Aust Econ Rev 2006; 39: 323–9. doi:10.1111/j.1467-8462. 2006.00422.x Schneider M-J. Introduction to public health. Burlington, MA: Jones & Bartlett Publishers; 2014. Australia HW. National cancer workforce strategy. Cancer workforce planning data inventory; 2011. Available at: http://www.hwa.gov.au/ sites/uploads/NCWS-Planning-Data-Inventory.pdf [verified November 2014]. Ljung L. System identification. Wiley Online Library; 1999. Available at: http://onlinelibrary.wiley.com/doi/10.1002/047134608X.W1046/ abstract [verified November 2014]. Kim S, Aurisicchio M, Wallace K. Towards automatic causality boundary identification from root cause analysis reports. J Intell Manuf 2009; 20: 581–91. doi:10.1007/s10845-008-0143-z D’Aveni R. Hypercompetition: managing the dynamics of strategic maneuvering. New York: The Free Press; 1994. Australia M. Medicare Benefits Schedule (MBS). Medicare, Australian Government, Department of Health; 2014. Available at: http://www. healthgovau/internet/mbsonline/publishingnsf/Content/Downloads-201 401 [verified 20 September 2014]. Randolph DS. Predicting the effect of extrinsic and intrinsic job satisfaction factors on recruitment and retention of rehabilitation professionals. J Healthc Manag 2005; 50: 49. Daniels ZM, VanLeit BJ, Skipper BJ, Sanders ML, Rhyne RL. Factors in recruiting and retaining health professionals for rural practice. J Rural Health 2007; 23: 62–71. doi:10.1111/j.1748-0361.2006.00 069.x Garrison G, Mikesell C, Matthew D. Medical school graduation and attrition rates. Analysis Brief 2007; 7: 2. Available at: https:// www.aamc.org/download/102346/data/aibvol7no2.pdf. [verified 20 November 2014]. Hayek FA. The use of knowledge in society. Am Econ Rev 1945; 35: 519–30. Dreyfus H, Dreyfus SE. Mind over machine. New York: The Free Press; 2000. Brooks PM, Lapsley HM, Butt DB. Medical workforce issues in Australia: tomorrow’s doctors: too few, too far. Med J Aust 2003; 179: 206–8. Australian Institute of Health and Welfare (AIHW). Health and community services labour force. National health labour force series. Number 42. Catalogue no. HWL 43; 2009. Available at: http://www. aihw.gov.au/WorkArea/DownloadAsset.aspx?id=6442458396 [verified 20 November 2014].

Australian Health Review

237

20 KPMG. Health workforce in Australia and factors for current shortages. Canberra: KPMG; 2009. Available at: http://www.ahwo.gov.au/docu ments/NHWT/The%20health%20workforce%20in%20Australia%20 and%20factors%20influencing%20current%20shortages.pdf [verified 20 November 2014]. 21 Becker GJ. 2000 RSNA annual oration in diagnostic radiology: the future of interventional radiology. Radiology 2001; 220: 281–92. doi:10. 1148/radiology.220.2.r01au39281 22 Pitman A, Jones D. Radiologist workloads in teaching hospital departments: measuring the workload. Australas Radiol 2006; 50: 12–20. doi:10.1111/j.1440-1673.2005.01524.x 23 Nakajima Y, Yamada K, Imamura K, Kobayashi K. Radiologist supply and workload: international comparison. Radiat Med 2008; 26: 455–65. doi:10.1007/s11604-008-0259-2 24 Jones N. Royal Australian and New Zealand College of Radiologists (RANZCR) radiology workforce survey 2000. Sydney: RANZCR; 2000. 25 Forrester JW. Industrial dynamics: a major breakthrough for decision makers. Harv Bus Rev 1958; 36: 37–66. 26 Richardson GP. Reflections on the foundations of system dynamics. Syst Dyn Rev 2011; 27: 219–43. doi:10.1002/sdr.462 27 Forrester JW. Urban dynamics. Cambridge: MIT Press; 1969. 28 Wolstenholme EF. System enquiry: a system dynamics approach. Chichester: John Wiley & Sons; 1990. 29 Forrester JW. Industrial dynamics. J Oper Res Soc 1997; 48: 1037–41. doi:10.1057/palgrave.jors.2600946 30 Angerhofer BJ, Angelides MC. System dynamics modelling in supply chain management: research review. Simulation Conference, 2000. Proceedings. Winter, vol.1, no., pp.342–351 vol.1, 2000. Available at: http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=899737&tag=1 [verified 20 November 2014]. doi:10.1109/WSC.2000.899737 31 Sterman JD. System dynamics modeling: tools for learning in a complex world. Calif Manage Rev 2001; 43: 8–25. doi:10.2307/41166098 32 Rodrigues A, Bowers J. The role of system dynamics in project management. Int J Proj Manag 1996; 14: 213–20. doi:10.1016/0263-7863 (95)00075-5 33 Brailsford SC, Desai SM, Viana J, editors. Towards the holy grail: combining system dynamics and discrete-event simulation in healthcare. Proceedings of the Simulation Conference, 2000; IEEE; 2010. 34 Brailsford SC, Evandrou M, Luff R, Shaw R, Viana J, Vlachantoni A, Willis RM . Using system dynamics to model the social care system: simulation modeling as the catalyst in linking demography to care delivery. Simulation Conference, 2000. Proceedings. Winter, vol.1, no., pp.342–351 vol.1, 2000. Available at: http://ieeexplore.ieee.org/xpls/ abs_all.jsp?arnumber=899737&tag=1 [verified 20 November 2014]. doi:10.1109/WSC.2000.899737 35 Samuel C, Gonapa K, Chaudhary P, Mishra A. Supply chain dynamics in healthcare services. Int J Health Care Qual Assur 2010; 23: 631–42. doi:10.1108/09526861011071562 36 Brailsford SC, Silverman E, Rossiter S, Bijak J, Shaw RJ, Viana J, Noble J, Efsathiou S, Vlachantoni A.. Complex systems modeling for supply and demand in health and social care.Simulation Conference (WSC), Proceedings of the 2011 Winter , vol., no., pp.1125–1136, 11–14 Dec. 2011. Available at: http://ieeexplore.ieee.org/stamp/stamp. jsp?tp=&arnumber=6147835&isnumber=6147014. doi:10.1109/WSC. 2011.6147835 37 Australian Bureau of Statistics 2012, Year Book Australia, 2012, cat. no. 1301.0, ABS, Canberra. Available at: http://www.abs.gov. au/ausstats/[email protected]/Lookup/by%20Subject/1301.0~2012~Main%20 Features~Population%20size%20and%20growth~47 38 Productivity Commission. Economic implications of an ageing Australia. Canberra: Productivity Commission; 2005. 39 Productivity Commission. Australia’s health workforce. Canberra: Productivity Commission; 2005.

238

Australian Health Review

40 Studdert DM, Mello MM, Brennan TA. Defensive medicine and tort reform: a wide view. J Gen Intern Med 2010; 25: 380–1. doi:10.1007/ s11606-010-1319-8 41 Meyboom-de Jong B. Feminiscring van de geneeskunde. Ned Tijdschr Geneeskd 1999; 143: 1134–5. 42 Jones VA. Why aren’t there more women surgeons? JAMA 2000; 283: 670. doi:10.1001/jama.283.5.670-JMS0202-5-1. 43 Phillips SP, Austin EB. The feminization of medicine and population health. JAMA 2009; 301: 863–4. doi:10.1001/jama.2009.155 44 Teljeur C, O’Dowd T. The feminisation of general practice: crisis or business as usual? Lancet 2009; 374: 1147. doi:10.1016/S0140-6736 (09)61742-6 45 Weizblit N, Noble J, Baerlocher MO. The feminisation of Canadian medicine and its impact upon doctor productivity. Med Educ 2009; 43: 442–8. doi:10.1111/j.1365-2923.2009.03309.x 46 Douglas A. The feminization of American culture. New York: Knopf; 1977. 47 Ferguson KE. The feminist case against bureaucracy. Philadelphia: Temple University Press; 1985. 48 Fondas N. Feminization unveiled: management qualities in contemporary writings. Acad Manage Rev 1997; 22: 257–82. 49 Burton KR, Wong IK. A force to contend with: the gender gap closes in Canadian medical schools. CMAJ 2004; 170: 1385–6. doi:10.1503/ cmaj.1040354 50 Barzansky B, Etzel SI. Medical schools in the United States. JAMA 2007; 298: 1071–7. doi:10.1001/jama.298.9.1071 51 Gould P. Data show shift toward more women in radiology departments. Diagn Imag Eur 2009; 25: 4–5. 52 Yelland CE, Yelland ME. Women in medicine: two generations. Med J Aust 2001; 174: 52–3. 53 Owen JB, Chan W, Sunshine J, Shaffer K. The sex ratio of American radiologists: comparison and implications by age, subspecialty, and type of practice. Am J Roentgenol 1995; 165: 1337–41. doi:10.2214/ajr. 165.6.7484559 54 Lewis RS, Bhargavan M, Sunshine JH. Women radiologists in the United States: results from the American College of Radiology’s 2003 survey. Radiology 2007; 242: 802–10. doi:10.1148/radiol.2423 060282 55 McKinstry B. Are there too many female medical graduates? Yes. BMJ 2008; 336: 748. doi:10.1136/bmj.39505.491065.94 56 Lambert TW, Goldacre MJ, Turner G. Career choices of United Kingdom medical graduates of 1999 and 2000: questionnaire surveys. BMJ 2003; 326: 194. doi:10.1136/bmj.326.7382.194 57 Australian Bureau of Statistics 1999, Labour Force Projections, Australia, 1999 - 2016, cat. no. 6260.0, ABS, Canberra. Available at: http://www.abs.gov.au/AUSSTATS/[email protected]/mf/6260.0 [verified 20 November 2014]. 58 Schofield DJ, Beard JR. Baby boomer doctors and nurses: demographic change and transitions to retirement. Med J Aust 2005; 183: 80. 59 Bhargavan M, Sunshine JH. Workload of radiologists in the United States in 2002–2003 and trends since 1991–19921. Radiology 2005; 236: 920–31. doi:10.1148/radiol.2363041316 60 May L, Martino S, McElveny C. The establishment of an advanced clinical role for radiographers in the United States. Radiography 2008; 14: e24–7. doi:10.1016/j.radi.2008.04.007 61 Bhargavan M, Sunshine JH, Schepps B. Too few radiologists? AJR Am J Roentgenol 2002; 178: 1075–82. doi:10.2214/ajr.178.5.1781075 62 First Consulting Group. The healthcare workforce shortage and its implications for American hospitals. Long Beach, CA: American Hospital Association; 2001. 63 Australian Institute of Health and Welfare (AIHW). Australia Health 2008: Section 8 – Expenditure and Workforce. AIHW, Canberra; 2009.

S. T. Taba et al.

64 Smith TN, Baird M. Radiographers’ role in radiological reporting: a model to support future demand. Med J Aust 2007; 186: 629–31. 65 Manning D, Ethell S, Donovan T, Crawford T. How do radiologists do it? The influence of experience and training on searching for chest nodules. Radiography 2006; 12: 134–42. doi:10.1016/j.radi.2005. 02.003 66 Brealey S, Piper K, King D, Bland M, Caddick J, Campbell P, Gibbon A, Highland A, Jenkins N, Petty D, Warren D. Observer agreement in the reporting of knee and lumbar spine magnetic resonance (MR) imaging examinations: selectively trained MR radiographers and consultant radiologists compared with an index radiologist. Eur J Radiol 2013; 82: e597–605. doi:10.1016/j.ejrad.2013.05.024 67 Brealey S, Scally A, Hahn S, Thomas N, Godfrey C, Coomarasamy A. Accuracy of radiographer plain radiograph reporting in clinical practice: a meta-analysis. Clin Radiol 2005; 60: 232–41. doi:10.1016/j.crad.2004. 07.012 68 Thompson W, Pollard K. The current status of radiographers as screen readers in breast screening units: applications for Australia. Radiograpgher 2007; 54: 16–20. 69 Freckelton I. Advanced practice in radiography and radiation therapy: report from the Inter-professional Advisory Team. 2012. Available at: http://www.air.asn.au/cms_files/09_AdvancedPractice/ipat_final_report 21april2012.pdf [verified 20 November 2014]. 70 Brown N, Leschke P.. Evaluating the true clinical utility of the red dot system in radiograph interpretation. J Med Image Radiat Oncol 2012; 56: 510–13. doi:10.1111/j.1754-9485.2012.02398.x 71 Advanced Practice Working Group (APWG). A model of advanced practice in medical imaging and radiation therapy in Australia. Melbourne: APWG; 2009. Available at: http://www.air.asn.au/advan ced.php [verified 30 October 2012]. 72 Smith T, Lewis S. Opportunities for role development for medical imaging practitioners in Australia: part 1. Rationale and potential. Radiographer 2002; 49: 161. 73 Swinburne K. Pattern recognition for radiographers. Lancet 1971; 297: 589–90. doi:10.1016/S0140-6736(71)91180-9 74 Davis LE. The coming crisis for production management: technology and organization. Int J Prod Res 1971; 9: 65–82. doi:10.1080/00207 547108929862  75 Ono T. Toyota production system: beyond large-scale production. Cambridge: Productivity Press; 1988. 76 Weick KE. The social psychology of organizing (topics in social psychology series). New York: McGraw-Hill; 1979. 77 Modis T. Fractal aspects of natural growth. Technol Forecast Soc Change 1994; 47: 63–73. doi:10.1016/0040-1625(94)90040-X 78 Schumpeter JA. Capitalism, socialism, and democracy. New York: Harper Perennial Modern Classics; 2008. 79 Mintzberg H, Jorgensen J, Dougherty D, Westley F. Some surprising things about collaboration: knowing how people connect makes it work better. Organ Dyn 1996; 25: 60–71. doi:10.1016/S0090-2616(96) 90041-8 80 Warren RB, Warren DI. The neighborhood organizer’s handbook. Notre Dame, IN: University of Notre Dame Press; 1977. 81 Backman EV, Smith SR. Healthy organizations, unhealthy communities? Nonprofit Manag Leadersh 2000; 10: 355–73. doi:10.1002/nml.10402 82 Kim J, Bentley PJ, editors. Immune memory in the dynamic clonal selection algorithm. Proceedings of the First International Conference on Artificial Immune Systems ICARIS; 2002: Citeseer. Available at: http://www.dtic.mil/dtic/tr/fulltext/u2/a474399.pdf#page=59 [verified 20 November 2014]. 83 Rawashdeh MA, Lee WB, Bourne RM, Ryan EA, Pietrzyk MW, Reed WM, Heard RC, Black DA, Brennan PC. Markers of good performance in mammography depend on number of annual readings. Radiology 2013; 269: 61–7. doi:10.1148/radiol.13122581

A systems approach to managing workforce

Australian Health Review

84 Reed WM, Lee WB, Cawson JN, Brennan PC. Malignancy detection in digital mammograms: important reader characteristics and required case numbers. Acad Radiol 2010; 17: 1409–13. doi:10.1016/j.acra. 2010.06.016 85 Jerslid S. Radiologist sightings drop round the world. Diagn Imaging (San Franc) 2003. Available at: http://www.diagnosticima ging.com/articles/radiologist-sightings-drop-around-world [verified 20 November 2014]. 86 Medical Radiation Practice Board of Australia. Professional capabilities for medical radiation practice. 2013. Available at: http://www.medical radiationpracticeboard.gov.au/Registration/Professional-Capabilities. aspx [verifed 20 November 2014]. 87 Adan I, Vissers J. Patient mix optimisation in hospital admission planning: a case study. Int J Oper Prod Manage 2002; 22: 445–61. doi:10.1108/01443570210420430

239

88 Cayirli T, Veral E. Outpatient scheduling in health care: a review of literature. Prod Oper Manag 2003; 12: 519–49. doi:10.1111/j.19375956.2003.tb00218.x 89 Cardoen B, Demeulemeester E, Beliën J. Operating room planning and scheduling: a literature review. Eur J Oper Res 2010; 201: 921–32. doi:10.1016/j.ejor.2009.04.011 90 Walter S. A comparison of appointment schedules in a hospital radiology department. Br J Prev Soc Med 1973; 27: 160–7. 91 Huang Y-L, Marcak J. Radiology scheduling with consideration of patient characteristics to improve patient access to care and medical resource utilization. Health Syst 2013; 2: 93–102.

www.publish.csiro.au/journals/ahr