Aotearoa New Zealand’s neurologist workforce: a 2024 analysis of demand, supply and projections

Abstract

Aims

To describe the Aotearoa New Zealand (NZ) neurology workforce, estimate neurology capacity, trends over time and future projections to assist with health sector planning.

Methods

2024 neurology workforce demographics, capacity and productivity were derived from multiple overlapping data sources including a nationwide individual neurologist and public hospital neurology unit head surveys, Health NZ and Royal Australasian College of Physicians data, and annual survey data from the Australian and New Zealand Association of Neurologists. Neurological service demand was assessed using epidemiological data, prior reported demand models and considered local practice patterns. Comparisons were made to our 2014 NZ report, Australian and international data. Projections were made modelling several scenarios.

Results

In 2024, there were 83 individual neurologists in NZ providing 67.3 full-time equivalents (FTEs) including 52.6 public adult, 8.3 paediatric and 6.4 private neurology FTE. The public sector FTE compared with 36 FTE in 2014. The combined adult headcount equates to 1 neurologist per 74 604 people (1.34/100 000) compared with Australia where there is 1 neurologist per 41 000 people (2.77/100 000) and 1 in 14 000 (7.1/100 000) in other high-income countries around the globe. A recent Australian workforce model estimated a minimum demand of 1 neurologist per 28 000 based on updated neurological disease and best-practice management requirements. If current training, recruitment, retention and practice patterns persist, projections indicate there will be a gradual worsening in the NZ neurology workforce over time.

Conclusions

The NZ neurology workforce has increased over the past 10 years, but NZ ranks well below other high-income countries globally, and demand continues to significantly exceed supply. This remains true, even after adjusting for NZ practice patterns which favour generalist over subspecialist care. If current trends continue, the gap is expected to widen rather than narrow over the next 12 years.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• We previously reported shortages in the New Zealand neurological workforce.

WHAT THIS STUDY ADDS

• This study provides a 10-year expanded update including time trends, paediatric, private and demographic data. The per population rate of neurologists has improved, but projections indicate a widening of the demand/supply gap by 2038, a persistent significant shortfall compared with other countries and a demography that does not mirror society.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• This study will support practice and policy efforts to diversify the workforce, increase neurologist positions, increase the number of neurology trainees, medical students and to concurrently explore novel technologies to enhance productivity.

Introduction

The worldwide burden of neurological disease is significant and rising.^{1 2} In 2021, conditions affecting the nervous system were collectively ranked as the leading group cause of disability adjusted life years (DALYs) (443 million, 95% UI 378 to 521), affecting 3.40 billion (3.20–3.62) individuals making up 43.1% (40.5–45.9%) of the global population. Global DALY counts attributed to these conditions increased by 18.2% (8.7–26.7%) between 1990 and 2021.¹

Worldwide, there has been a longstanding underprovision of neurological services for a variety of reasons. This gap in service provision is forecast to widen and the pressures on the health system are going to increase over the next decades.²

In 2014, we published Aotearoa New Zealand (NZ) neurological workforce data exploring several models to help close the demand and supply gap that went beyond increasing the number of neurologists to also consider opportunities for efficiency gains through technology and use of alternative workforces.³

This report provides an update on the evolving NZ neurologist workforce. To ensure comparability, we used a similar methodology to estimate supply and demand to what we used in 2014.³ Other models are also presented, including one from a recent Australian study, to allow comparisons to other published data.⁴ For the first time, we also present NZ neurology workforce extended demographic data including paediatric and private neurologist data. The overall aim is to assist with health workforce planning.

Methods

This study considers public and private practice NZ specialist neurologist workforce with a focus on adult neurology but includes paediatric neurologist demographic data. We attempted to capture the contribution of non-medical service providers such as neurophysiology technicians, neurology nurses and neurology trainees, but data was inconsistent and had to be excluded from the final analysis. Their contributions are captured in the overall service output and productivity.

Neurologist demographic data

In May 2024, we emailed a survey to all NZ-based neurologists to identify their current demography. Neurologists were identified via the Neurological Association of New Zealand membership email list supplemented with additional emails to personal contacts. For cross-validation purposes, we compared both head and full-time equivalent (FTE) counts against additional survey data contributed by all public hospital unit heads, data from Health NZ, the Australian and New Zealand Association of Neurologists (ANZAN) and the Royal Australasian College of Physicians (RACP).

Requested information included age, ethnicity, gender, sexual identity, place of training, current practice location, public, private, academic, leadership and other roles and associated FTE, subspecialisation and anticipated retirement age. When determining FTE, clinical and related administrative and leadership time was counted. Academic activity and other non-clinical activities such as hospital level leadership roles were excluded.

Workforce model

The demand for specialist neurological services depends primarily on the incidence and prevalence of neurological disease and is subject to population size and demographic change. Other contributing factors include public awareness of neurological disease, public health interventions, availability of diagnostic and therapeutic options, quality of primary care services and care complexity, referral and service protocols, and the availability of non-neurology specialists (eg, geriatricians and internal medicine specialists) to manage some of the common conditions (figure 1).

Figure 1. Model of neurological service demand and supply.

The supply of the specialist neurological workforce depends on the number of funded public neurologist positions, additional private neurologists, sufficient NZ neurology trainees to fill positions, ability to recruit and retain both NZ and overseas trained neurologists, and retirement. Finally, advanced practice providers, use of technology and efficient models of care may positively impact productivity and thus ‘supply’ if these are factored into the model.³

Estimating supply and productivity

To assess the current supply of neurological services, we conducted a unit head survey to attain FTEs of specialist neurologists and neurologist head counts in each of the public hospitals in NZ. To determine per FTE productivity, we considered annual patient volumes and average time spent on each specific clinical activity. We validated this data against the individual demographic survey, Health NZ, ANZAN and RACP data.

Neurologists perform a range of clinical activities of varying duration and complexity. To account for these variations, we employed the same methodology as in our 2014 report.³ Briefly, we converted all reported public sector adult neurology case volumes to a proportion of a first specialist assessment (FSA or new neurology assessment), called patient contact equivalent (PCE), based on average consultation duration reported by unit heads for all counted activities. Counted neurologist activities included FSAs, follow-ups (FUs), botulinum toxin injections, electromyography (EMG), inpatient admissions and consultations, repeat ward round encounters, telehealth/telestroke assessments, regional referral telephone calls and non-contact assessments (‘virtual clinics’). PCEs do not reflect the total workload of a specialist neurologist but allow allocation of weights to all countable patient encounters. This allows the estimation of an annual PCE figure that one full-time neurologist can provide. Dividing the total number of PCEs per year by currently available neurologist FTEs provides a figure of cases/FTE to estimate current productivity level.

$${\displaystyle {\displaystyle \mathrm{T}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}\mathrm{n}\mathrm{u}\mathrm{m}\mathrm{b}\mathrm{e}\mathrm{r}\mathrm{o}\mathrm{f}\mathrm{a}\mathrm{n}\mathrm{n}\mathrm{u}\mathrm{a}\mathrm{l}\mathrm{P}\mathrm{C}\mathrm{E}\mathrm{s}/\mathrm{T}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}\mathrm{n}\mathrm{u}\mathrm{m}\mathrm{b}\mathrm{e}\mathrm{r}\mathrm{o}\mathrm{f}\mathrm{F}\mathrm{T}\mathrm{E}\mathrm{s}=\mathrm{P}\mathrm{r}\mathrm{o}\mathrm{d}\mathrm{u}\mathrm{c}\mathrm{t}\mathrm{i}\mathrm{v}\mathrm{i}\mathrm{t}\mathrm{y}}}$$

This productivity figure represents an average across neurology departments in the country. Activities such as reporting diagnostic tests, administrative tasks, teaching, leadership and many other activities that neurologists routinely engage in were considered when arriving at the FTE supply estimates but are not numerically reflected in the PCE figure. This figure also averages varying degrees of sub-specialisation, models of service and referral pathways and incorporates contributions from other departmental staff. We considered anticipated retirement rates and current training positions. The analysis does not consider immigration or emigration, assuming these flows will remain relatively stable over time.

Estimating demand

NZ epidemiological data for neurological disease is limited and we primarily estimated incidence and prevalence rates using international evidence,^{5 6} considering NZ data where available.^{5 7}

To allow comparison to our 2014 study, the base model used the same assumptions and data sources but adjusted for interim population growth.³ Our prior assumptions incorporated an adjustment for a NZ neurology practice pattern that tends to focus on tertiary level neurology with much of primary and secondary neurology managed by general practitioners (GPs), general physicians and geriatricians.

After adjusting international figures to fit the NZ landscape, incidence and prevalence rate estimates suggest that annually 0.6% of the population requires an FSA and 3.6% of the population should be under FU care of a neurologist at any one time. However, NZ health resources are constrained compared with other high-income countries. A strong political push for timely FSA assessments in a resource-limited environment has resulted in a specialist FU appointment rate far lower than expected based on disease prevalence. Based on incidence and prevalence, we should expect about six times as many FU appointments as FSAs; however, Health NZ reports an overall ratio of 1:1 FSA to FU (personal communication). This means patients are generally discharged back to their GP after a single visit. Only about one in five patients with chronic neurological disease is regularly reviewed by a neurologist, and a proportion of patients is referred for a second ‘FSA’ after some time.

To account for this, our 2014 report proposed an average demand of 1.2 PCEs per year for incident and 0.2 PCEs for prevalent cases. The total estimated demand for PCEs per year was then calculated by adding total demand for FSAs (estimated incidence multiplied by 1.2 PCEs) and demand for FU consults (prevalence multiplied by 0.2 PCEs).

We then multiplied the volumes of required PCEs by the average PCE-based productivity per one FTE neurologist to arrive at a demand figure displayed as the total consultant neurologist FTE. Finally, we projected this demand over the next 12 years based on population growth estimates.⁸ These figures were not adjusted for changes relating to population age and ethnic distribution. Such adjustments would likely further increase estimated demand, making our projections conservative. In addition to the baseline model, we also considered changes in productivity since 2014 and modelled neurologist demand over the next 12 years taking these trends into account.

Benchmarking supply and demand estimates

The above methodology provides an estimate of current demand incorporating epidemiological and productivity data within the NZ context against current neurologist FTE supply. This provides a gap analysis along with a projection over the next 12 years.

For the purposes of benchmarking, we present 12-year projections using our baseline model, productivity adjusted model and neurologist FTE requirements based on published international demand estimates.^{4 6}

Results

Estimated supply

We identified a total of 83 neurologists via the individual and unit head surveys with an associated 67.3 FTE, that is, an average FTE of 0.81 per neurologist. This overall figure includes 52.6 public adult FTE, 8.3 public paediatric FTE and 6.4 private FTE. Of the 83 identified individuals, 54 (65.0%) completed an individual survey providing demographic data. Some respondents did not complete all fields or entered uninterpretable data. For example, for ethnicity or sexual identity, several respondents reported ‘New Zealander’ or ‘male/female’, respectively; thus, responses had to be excluded, likely skewing results (table 1).

Table 1. Neurologist demographics.

	All (including n=3 paediatrics) N=54
		Response, n
Age, mean (SD) years	52 (12.3)	46 (85)
Female, n (%)	15 (32.6)	46 (85)
LGBTQI+, n (%)	4 (20)*	20 (37)
Ethnicity
NZ and other European  Māori  Asian	34 (79.1)† 3 (7.0) 6 (14.0)	43 (80)
Years in NZ practice, mean (SD)	14.2 (11.7)	46 (85)
Future years in NZ neurology practice, mean (SD)	14.2 (9.9)	40 (75)
Planned retirement age, mean (range)	65 (52–81)	40 (75)
Primary work city, n (%)‡
Auckland  Wellington  Christchurch  Dunedin  Palmerston North  Hamilton  Tauranga  Hastings  Nelson  Rotorua  Gisborne	15 (29.4) 10 (19.6) 7 (13.7) 5 (9.8) 4 (7.8) 3 (5.9) 2 (3.9) 2 (3.9) 1 (2.0) 1 (2.0) 1 (2.0)	51 (94)
Primary training country, n (%)
NZ  UK  Europe  South Africa  North America  India	27 (64.3) 6 (14.3) 4 (9.5) 3 (7.1) 1 (2.4) 1 (2.4)	42 (78)
FTE, mean (SD)
Public  Private  Leadership  Academic  Total	0.78 (0.31) 0.25 (0.2.0) 0.13 (0.06) 0.33 (0.27) 0.94 (0.42)	48 (89) 26 (48) 9 (17) 14 (26) 51 (94)
Primary subspecialty interest
MS/neuroimmunology  Stroke  Epilepsy  Neuromuscular  General neurology  Movement disorders  Neurogenetics  Cognitive/behavioural  Headache  Vestibular  Clinical neurophysiology  Medicolegal	10 (21.3) 9 (19.2) 8 (17.0) 7 (14.9) 5 (10.6) 2 (4.3) 2 (4.3) 1 (2.1) 1 (2.1) 1 (2.1) 1 (2.1) 1 (2.1)	47 (87)
Time spent in subspecialty, % (range)	26.5 (0–100)	47 (87)

^*n=20 responded, however, n=9 simply restated their sex, which was interpreted as ‘non-LGBTQI+’; of n=11 with appropriate response n=4 self-identified as LGBTQI+ (ie, 36%).

^†All Māori self-identified also as NZ European; n=3 excluded as reported ‘New Zealander’ (n=2) or ‘N/A’ (n=1) as these are not considered ethnicities as per the NZ official demographic reporting convention.

^‡Other Cities covered via outreach services: South Auckland, North Shore, Whangarei, Invercargill, Balclutha, Dunstan, New Plymouth, Greymouth, Wairoa, Taupo, Hutt, Masterton, Porirua, Nelson, Napier.

FTE, full-time equivalent; MS, Multiple Sclerosis; NZ, New Zealand.

We explored demographic data by adult versus paediatric and/or excluding private only (n=3) and low (<0.2) FTE individuals (n=3). Overall, patterns did not differ substantially with the exception of paediatric neurologists being almost exclusively located in tertiary centres. We do not present further disaggregated data given small practitioner volumes and the associated risk of identification.

We converted the above FTE to neurologists per population by dividing total FTE by the 2024 NZ population, resulting in 1 neurologist per 77 563 people, with 1/88 469 for all adult neurologists and 1/99 227 adult neurologists in public hospitals. Comparing this to the 2014 study, which only considered public adult neurologists, this represents a 46.1% absolute increase in public FTE (36 in 2024 vs 52.6 in 2024) and a 27.3% proportional improvement from 1/126 000 to 1/99 227 population.

Based on reported retirement plans, we estimate a reduction of 1.7 neurologists per year over the next 12 years. This is offset by up to five neurology advanced trainees (‘trainees’) completing their training each year. However, with only 1.7 vacancies created due to retirement, 3.3 will be required to work in the private sector or move overseas unless there is an increase in publicly funded positions. We have observed growth of 1.3 additional public neurology positions per year over the past 12 years. If this continues, two out of five trainees will still need to find alternatives.

From a geographical perspective, 53.1 of the 67.3 total neurologist FTE (78.9%) work at tertiary or metropolitan centres (Auckland, Christchurch, Wellington, Hamilton and Dunedin) and 14.2 FTE (21.1%) in regional secondary settings. This translates into 1/70 267 total neurologists in metropolitan and/or tertiary and 1/104 859 in non-metro/tertiary centres.

Internationally, medical workforce data is sometimes presented as doctor head count/population or head count per 100 000. To allow comparisons, our total neurologist head count corresponds to 1/62 919 (or 1.58/100 000); 1/71 538 (1.39/100 000) for adult public and private, and 1/74 604 (1.34/100 000) for adult public alone.

Estimated demand

Table 2 presents current and 12-year projected total NZ population volumes, estimated population of New Zealanders with incident and prevalent neurological care, and the conversion to annual PCEs required to meet this need based on current NZ practice patterns.

Table 2. Neurology specialist service demand in NZ over the next 12 years.

Year (Projected total NZ population)	Incident (new) cases Prevalent (existing) cases	PCEs /year*	Estimated demand for neurology specialists in NZ
			A NZ (current productivity)	B NZ (lower productivity)	C 1/50K (Kurztke needed)	D 1/41K (Australia current)	E 1/32K (WFN current)	F 1/28K (Australia needed)
2024 (4 511 400)	32 140 192 841	77 136	98	98	107	131	167	191
2025 (4 553 280)	32 433 194 596	77 838	98	99	108	132	169	193
2026 (4 595 450)	32 728 196 367	78 547	99	101	109	133	170	195
2027 (637 390)	33 026 198 154	79 262	100	103	110	134	172	197
2028 (4 679 585)	33 326 199 957	79 983	101	105	111	135	174	198
2029 (4 721 465)	33 629 201 777	80 711	102	107	112	137	175	200
2030 (4 762 765)	33 935 203 613	81 445	103	109	113	138	177	202
2031 (4 804 050)	34 244 205 466	82 186	104	111	114	139	178	204
2032 (4 845 520)	34 556 207 335	82 934	105	114	115	140	180	206
2033 (4 887 325)	34 870 209 222	83 689	106	116	116	142	182	208
2034 (4 928 175)	35 188 211 126	84 450	107	118	117	143	183	209
2035 (4 968 660)	35 508 213 047	85 219	108	120	118	144	185	211
2036 (5 008 605)	35 831 214 986	85 994	109	123	119	146	187	213

A: estimated number of specialists required in NZ=total PCEs required in a year/PCEs provided per specialist at current productivity.

B: estimated number of specialists required in NZ at reducing productivity of 10% reduction over next 12 years PCEs/specialist/year.

C: estimated number of specialists required in NZ at the specialist population ratio of 1/50 000 as suggested by Kurtzke in the USA.

D: estimated number of specialists required in NZ at the specialist population ratio of 1/41 000 as reported to be in place in Australia in 2020.

E: estimated number of specialists required in NZ at the specialist population ratio of 1/32 000 as suggested compared with WFN average globally.

F: estimated number of specialists required in NZ at the specialist population ratio of 1/28 000 as suggested by Australian model.

^*Estimated total number of PCEs /year = (estimated Incidence × 1.2 PCEs/case/year) + (estimated prevalence × 0.2 PCEs/ case/year).

NZ, New Zealand; PCEs, patient contact equivalents; WFN, World Federation of Neurology.

We estimated the current average productivity rate of NZ neurologists is 791 PCEs/specialist/year based on the unit head surveys describing average duration of an FSA (46.5 min) and FU (26 min) appointment as well as reported current case volumes. Dividing the total yearly demand for PCEs by current productivity rate allowed a prediction of demand for FTE neurology specialists over the next 12 years if the current supply to demand ratio persists, taking into account population growth alone (table 2, Column A).

This analysis was compared with four additional scenarios. Column B in table 2 shows demand for neurology FTEs assuming a 10% productivity decrease based on the observation of a 10% PCE reduction of 875 to 791 between our 2014 and 2024 studies. The other three scenarios (Columns C, D, E and F) show predicted demand on the basis of specialist to population ratios in line with Kurtzke’s 1982 estimates (Column C),¹ the 2020 Australian study (Column D),⁴ the World Federation of Neurologists reporting current average number of neurologists in any country (1/32 000 Column E)² and the modelled actual need considering updated modern neurological demands in Australia (1/28 000; Column F). In the 2014 report, we included a conservative 2011 Royal College of Physicians UK model of 1:70 000⁹ and a 1989 NZ workforce report recommendation of an even more conservative estimate of 1:100 000.¹⁰ However, based on the Australian and World Federation of Neurology estimates, these figures now seem outdated and have been removed as a comparator.

Mapping supply against demand

Figure 2 presents supply versus demand projections and various models to bridge the current gap. In contrast to the 2014 report, we present all adult (private and public) data for this analysis to allow better benchmarking with other countries and to more accurately reflect adult neurologist availability in NZ. We retain primary presentation by FTE rather than head count to account for an increasing number of part-time neurologists over time.

Figure 2. Projected supply and demand for the neurology specialist workforce in NZ. Aus, Australia; NZ, New Zealand.

For our baseline supply model 1, we incorporated the current annual 1.3 neurologist workforce growth rate and productivity patterns without any additional changes. By 2036, this would result in a supply of 76.6 neurologist FTE against a projected demand of 98 FTE, a gap of 46.3 neurologist FTE compared with the current gap of 28.7. For supply model 2, we added an additional 3.2 (total of +4.5/year) neurologists annually required to meet projected NZ demand of 122.9 neurologists by 2036, and supply model 3 adds an additional 10.7 (total of +12/year) to meet the Australian projected demand of 213 neurologists by 2036. The Australian and NZ demand lines correspond to data depicted in table 2, columns B and F.

Data validity

To ensure maximal validity, we considered multiple overlapping data sources. Aside from the ANZAN survey, there was reasonable alignment across sources (figure 3).

Figure 3. Neurologist head count and FTE from various 2024 data sources. NZ, New Zealand.

Health NZ data only considers publicly funded FTE and lead author (AR) personally reviewed the Health NZ data with a representative removing incorrect data fields such as neurology fellows, indicating that all data sources require a degree of scrutiny. Furthermore, while the totals reasonably align, the variation by unit was extensive, with one hospital being reported to have anywhere between 6 and 25 neurologists depending on the data source.

Discussion

Our updated neurology workforce survey demonstrates a growth in the public adult neurology workforce from 1 FTE neurologists/126 000 to 1/99 227 people between 2014 and 2024. However, even the updated, more inclusive figure of 1/88 469 (59 FTE) that includes private adult neurology services still falls short of our conservative current demand estimates of 1/46 000 (98 FTE). A recent NZ cardiology workforce report identified a ‘deep crisis’ with 1 cardiologist/30 487 people.¹¹ Considering stroke alone is the second leading cause of death globally after ischaemic heart disease, treatment interventions are increasingly mimicking cardiology care and neurological disability is steadily rising due to the ageing population, this crisis is at least shared.¹

It is of interest to compare our NZ estimates to international models and rates of neurologists for benchmarking purposes. Kurtzke, considering the USA in 1982, estimated a ratio of 800 FSA and 1200 FU neurological consultations per year per 100 000 people, or approximately 1 neurologist/50 000 people.¹ The latest neurology workforce data available from the USA indicates that this target is being met, although with a wide geographical range between 1 neurologist/56 000 people in rural areas and 1/9000 people in metropolitan areas.¹² The World Federation of Neurology reported 1/32 000 neurologists/populations globally and 1/14 000 in high-income countries.

The more recent 2020 Australian neurology workforce analysis reported 620 neurologists with 1 neurologist per 41 000 people meeting Kurtzke’s demand estimates. Yet, the authors proposed a significant deficit modelled on the assumption that instead of Kurtzke’s 1:1.5 ratio, NZ’s 1:1 or Australia’s 1:2 of FSA to FU consultation per year, a ratio of 1:9 was more realistic, resulting in an estimated neurologists demand of 1/28 000.⁴

While the underlying incidence or prevalence of neurological diseases is unlikely to have changed dramatically over time, the increased availability of treatments is driving most of this change. For example, multiple sclerosis treatment options have become more complex, requiring specialist input and new Alzheimer’s treatments are on the horizon. Rarer diseases such as spinal muscular atrophy now have treatment options with many more treatments for neurogenetic diseases imminent. Other practice changes increasing neurologist demand include advances in acute stroke reperfusion treatments with more patients who had a stroke admitted under the care of neurologists and tertiary units now routinely providing telemedicine or telephone expert decision support to secondary centres and accepting many more patients in transfer for reperfusion therapy.

To achieve even the conservative NZ neurology demand projections of 1/46 000, current annual FTE increases would have to more than triple from 1.3 new FTE per year to 4.5 FTE per year. With a 3-year training programme and 13 training posts, we currently have the capacity to train a maximum of four to five new neurologists each year, offset by the 1.7 neurologists retiring each year. This means we need to increase trainee positions by at least 1.2 FTE and probably closer to 2–3 given anticipated international attrition. An additional challenge is presented by vacancies among existing training positions. To achieve Australian demand projections, we would have to increase annual trainee intake by at least nine trainees, requiring an increase in medical students.

Our demographic data found increasing female neurologists, with 32.6% females in 2024 compared with 23.3% in 2014 (unpublished). However, NZ still falls behind countries like the USA where 48.3% of neurologists are now female.¹³ Our results on ethnicity and gender identity are encouraging, but neurologist ethnicity proportions still do not match society. Clinical workforces that match the demographics of the populations they care for improve patient outcomes.¹⁴ Reporting on these results is important not only to monitor progress, but also to signal aspiring doctors and neurologists that the neurology community is welcoming of and values people from diverse backgrounds.

For the first time, we report neurology subspecialisation patterns in NZ, which demonstrates a potential mismatch between subspecialists in a given area with disease prevalence and clinical need. Further exploration of these patterns and a more deliberate approach to planning subspecialisation and recruitment may be warranted.

We found an encouraging 1:4 ratio of non-metro/non-tertiary versus metro/tertiary neurologist distribution in NZ compared with a 1:24 ratio in Australia. Yet, there is still ample opportunity to improve outreach to smaller centres and future analyses could include a more detailed analysis on geographical access to neurological services. Our surveys attempted to capture some of this information, but it was too inconsistently reported.

In 2014, we modelled various approaches aimed at increasing neurologist productivity through new models of care including telehealth and increasing trainee positions and neurology nurse positions. Surprisingly, despite reports that these approaches have been adopted, neurologist PCE productivity has decreased by 10% over the past 10 years. There are several possible explanations. First, the combination of increases in disease complexity and increased operation at ‘top of scope’ through better referral pathways may have resulted in a net reduction in per patient productivity. Had the new care models not been adopted, the current situation would likely be even worse. Second, both unscheduled acute and night-time activities may have disproportionately increased. This type of work is inherently less efficient and resource intensive and can produce a picture of reduced inefficiency while at the same time negatively impacting staff well-being and increasing risk of burnout. Third, some uncounted activities, such as calls from local GPs, administrative burden and quality assurance activities may have increased over time. Fourth, work patterns prioritising doctors’ wellness and improved work-life balance may have resulted in decreased productivity. Finally, an increasing number of nurses and trainees may be offset by additional supervision demands on neurologists, potentially resulting in better quality rather than higher quantity of care. Productivity requires further exploration both as regards potential barriers and identifying and testing solutions. For example, advancing artificial intelligence technologies to improve referral pathways, triaging, diagnosis and long-term management shows great promise¹⁵ and should be explored to complement a drive to increase the number of neurologists.

To address the reported workforce shortages, several key changes could be considered. First, it is imperative that the maximum number of graduating trainees is retained in NZ. To achieve this, jobs need to be attractive and available, with supporting business cases to health entities prioritised. Second, to achieve even the modest NZ-based demand projections, we need to increase trainee positions and implement strategies to attract more medical students into the field of neurology. Clinical observerships, summer studentships and ample exposure in medical curricula are important and readily justified given 22% of hospitalisations are neurological in nature.¹² There will also be opportunities to improve student and trainee experience within neurology departments.¹⁶ While it is important to continuously explore new models of care and alternate workforces, it remains uncertain how much impact these efforts can have on specialist workforce demands, and more data is needed.

This study has several strengths. The use of multiple overlapping data sources increases confidence in reported results. Exploring and presenting multiple demand models provides broader sensitivity and confidence margins presenting a range of potential targets to aim for and acknowledges that many of the considerations are based on assumptions. The 100% unit head completion rate and overall high completion rate of individual surveys is an excellent achievement and attributable to the relatively small and tight-knit NZ community of highly committed neurologists that operate in a generally supportive and collegial environment.

This study also has limitations. While the adult public FTE and head count data is likely accurate, we had some incomplete data from the paediatric neurology community, and private practice data is difficult to validate. It would have been preferable to specifically and directly consider non-neurologist clinicians’ contribution to neurology service supply and future work should consider novel approaches to achieve this. In addition, the NZ demand model includes several assumptions, some of which have not been fully tested. We attempted to address this by presenting alternative international models for benchmarking. Finally, different models around the globe use different variables in workforce reports. We focused on FTE while others used head count, which we speculate is mostly used because this data is more readily accessible. We maintain that an FTE approach is more meaningful, especially as the workforce changes with an increasing number of part-time neurologists. However, to allow for wider comparison, we also presented head count data.

In conclusion, the NZ neurologist workforce is growing, slowly moving toward reflecting the NZ population and comparing favourably when it comes to metro/tertiary to non-metro/non-tertiary distribution. However, compared with Australian and World Federation of Neurology figures, our number of neurologists falls woefully short and does not meet even the most conservative demand figures. NZ requires strong funding, recruitment and training initiatives if we want to be ready for the projected increase in neurological burden of disease now and over the next decade.

Data availability statement

Data are available upon reasonable request.