Holding a mirror to society? The socio-demographic characteristics of students commencing health professional programmes, and all courses, at Ōtākou Whakaihu Waka (the University of Otago), 1994–2023

Academic institutions play a pivotal role in shaping the health workforce via the effects of their policies and practices on recruitment and retention of students.^1,2 Ōtākou Whakaihu Waka (the University of Otago) recognises its responsibility to develop a health workforce that is equipped to meet the needs of society, and strives to ensure that health professional programme student cohorts reflect the varied ethnic, socio-economic and geographic contexts of Aotearoa New Zealand (Aotearoa) communities.³

“Affirmative action” refers to strategies that aim to increase representation of groups that have historically been under-represented or excluded, and is undertaken in health professional education with a number of Te Tiriti o Waitangi–based and equity-based justifications.⁴ These include honouring the Crown’s obligations under Te Tiriti o Waitangi (in Aotearoa specifically); correcting current and historical injustices; the benefits that concordant backgrounds have for patient–practitioner relationships; and the observation that practitioners from under-represented communities are more likely to return to serve those communities.^4–9 Although initially limited and with major shortcomings, affirmative action admission policies have existed at Ōtākou Whakaihu Waka since at least as early as 1951.^10,11 The university’s current policy, Te Kauae Parāoa,³ applies to eight health professional programmes and aims to facilitate the entry of students who are Māori, Indigenous Pacific, from rural backgrounds, from refugee backgrounds or who studied at schools that serve communities with higher socio-economic disadvantage.

With these measures in place, it is important to evaluate what changes have occurred in the composition of student cohorts. Cross-sectional analyses of the socio-demographic profile of students entering health professional programmes at Ōtākou Whakaihu Waka were previously undertaken for the years 2010 and 2016.^12,13 Recently we undertook more extensive analyses, using automated processes to produce a longitudinal socio-demographic “atlas” of students entering all courses at Ōtākou Whakaihu Waka, all health professional programmes combined and 11 individual health professional programmes between 1994 and 2023 (see supplementary material). This research was initiated by academics in the university as a further contribution to monitoring student participation, and was carried out in collaboration with the University of Otago’s Strategy, Analytics and Reporting Office. This paper presents an overview and discussion of selected key findings.

Methods

This section summarises key aspects of methods that are relevant to findings described in this article. Full methods are outlined in the accompanying report (see supplementary material, p. 5).

University data

Ōtākou Whakaihu Waka maintains electronic collections of student data extending back to 1993, covering programme details as well as demographics, information on schooling and home addresses. Ethnicity is self-identified within university records, with students currently being able to select up to three different groups. For most of the study period, collection of gender information has been limited to a binary male/female classification; students have only been able to self-select that they identify with another gender in recent years. Home addresses used for this study were the addresses that students provided at their first-ever enrolment with the university. Commencement data for 1993 include all students who were enrolled at any level in the university during that year; therefore, commencement dates for students enrolled in 1993 are not reliable.

Derivation of study cohorts

Enrolment data were extracted from university records from 6 to 11 October 2023 and were used to derive two cohorts. The first cohort comprised all students who commenced health professional programmes offered by Ōtākou Whakaihu Waka between 1994 and 2023 that would normally lead to registration under the Health Practitioners Competence Assurance Act 2003¹⁴ (Table 1). Students were included once for each health professional programme they fully enrolled in, such that students who commenced more than one health professional programme over time were included for each instance.

View Table 1–4, Figure 1–3.

The second cohort comprised all students who commenced any programme at Ōtākou Whakaihu Waka between 1994 and 2023. Inclusion in this cohort aimed to approximate students coming to the university for the first time or returning to start a new episode of study after a substantial period away. Therefore, students were included for any discrete instance where they 1) commenced a new programme of study not previously enrolled in; and 2) had no other enrolment recorded at Ōtākou Whakaihu Waka during the previous two academic years and; 3) were not already included due to enrolment in another programme in the same year.

Identification of students commencing health professional programmes was undertaken independently of identification of students commencing all programmes university-wide.

Key variables

Socio-demographic variables

Socio-demographic variables were derived and classified using information present in enrolment data and linkage to other publicly available datasets.^15–18

Residency status was classified based on students’ recorded residencies at the time of programme commencement. Students who were not New Zealand citizens or permanent residents were excluded from most analyses.

Ethnic group was classified from the most recent available university records using the Statistics New Zealand (Stats NZ) level one categories.¹⁹ For most analyses, prioritised output was used (whereby individuals identifying with multiple ethnic groups are assigned a single ethnic group using an order of priority). The groups used (in descending order of priority) were: Māori; Pacific peoples; Asian; Middle Eastern/Latin American/African (MELAA); and European or Other. For calculation of rates, the same groups were used but ethnicity was classified using total response output (where individuals identifying with multiple ethnic groups are included in totals for each), to match available population denominators.

Age group was classified based on students’ ages on 1 March in the year of commencement. Gender was classified based on the Stats NZ standard output for gender²⁰ (male, female or another gender) using the most recent available university records.

Socio-economic deprivation was classified as the New Zealand small-area index of relative socio-economic deprivation (New Zealand Index of Deprivation [NZDep])²¹ quintile of students’ home addresses. NZDep is an index of relative socio-economic deprivation produced for small areas using aggregated information from census data. NZDep versions have been produced for each census year since 1991.¹⁶ The “tidygeocoder” package²² in R was used to pass addresses to the ArcGIS geocoding service, which identified the latitude and longitude of students’ home addresses that in turn enabled matching of those addresses to the relevant meshblocks.

School socio-economic quintiles were derived from the decile of students’ last-attended schools.¹⁸ School deciles were used by the Ministry of Education between 1995 and 2022 (after which they were replaced by the new Equity Index system), and were derived in each census year from the socio-economic makeup of attending students’ neighbourhoods.²³ Decile 1 schools were the 10% of schools with the highest proportion of students from low socio-economic communities, and decile 10 schools were the 10% of schools with the lowest proportion of these students. A school decile did not measure the standard of education delivered at a school. In most instances, NZDep and school deciles were applied such that the version that most closely approximated students’ year of programme commencement was used (see supplementary material, p. 12, for more detail).

Urban/rural status was defined using the 2018 Geographic Classification for Health (GCH)^17,24 of students’ home addresses. The GCH uses a combination of census-derived urban area population counts and travel time to the edge of urban areas to classify small areas into five levels of urbanicity/rurality.²⁴ For most analyses, the GCH was determined at Statistical Area 1 level; however, for calculation of rates, Statistical Area 2 classifications were used to match available population denominators.

Admission details

Admission categories and subcategories were available for most health professional programmes from 2017 onwards. Admission categories are based on the amount and nature of study previously completed by applicants (e.g., secondary school, health sciences first year, bachelor-level qualification, alternative [health-related professional experience in a relevant field]). Admission subcategories are open to students at all levels of study, and include affirmative action pathways as well as the international subcategories. General category applicants are those who do not apply via a subcategory pathway.

Analyses

Most analyses in this article present simple counts and proportions of commencing students, disaggregated by time period (where applicable) and socio-demographic variables. For longitudinal analyses of students commencing by 5-year time period (Figure 1–3), the mean number commencing per year in each period was plotted. Small numbers (<5 students) were not reported in order to protect privacy.

Analyses of commencement rates per 100,000 population (undertaken for ethnic group, GCH and region) were restricted to students aged 18–29 years and used populations aged 18–29 years as the denominator to account for differences in underlying population age structures. For ethnicity rate calculations, total response output was used to match available population denominators. Population denominators were derived from publicly available²⁵ and bespoke estimates of Aotearoa sub-populations over the study period (Stats NZ, customised report and licensed by Stats NZ for re-use under the Creative Commons Attribution 4.0 International licence). See supplementary material, p. 15, for full details on rate calculations.

Software and automation

These analyses were undertaken primarily using R version 4.3.2.²⁶ To produce the breadth of included analyses (see supplementary report, p. 17), an automated approach was taken whereby functions were developed to undertake all calculations across all variables and programmes and produce formatted tables and figures. A template was developed using the document writing package RMarkdown²⁷ that applied these functions to write “chapters” for each programme category (e.g., “All Health Professional Programmes”, “Bachelor of Physiotherapy”), containing all desired statistical output. The template was run in sequence across each of the 13 included programme categories to produce the results section of the full report (supplementary material, p. 17).

Ethics statement

This project received ethical approval from the Ōtākou Whakaihu Waka Human Ethics Committee (reference number D23/277).

Results

This section presents an overview of key findings, with a focus on students commencing all courses at Ōtākou Whakaihu Waka, all health professional programmes combined and the Bachelor of Medicine and Bachelor of Surgery. The Bachelor of Medicine and Bachelor of Surgery is highlighted as it is the largest programme, and the programme for which affirmative action policies have had the most influence. Comprehensive statistical summaries for each programme category including analyses of age, regional origins and programme completion (for health professional programmes) are in the full report (supplementary material).

Total numbers and residency status

This study identified 182,932 records for students commencing new episodes of study in all programmes at Ōtākou Whakaihu Waka between 1994 and 2023, of whom 148,653 (81.3%) were Aotearoa citizens or permanent residents. Over that same period, 20,978 records were identified for students who newly commenced health professional programmes, including 18,632 (88.8%) who were Aotearoa citizens or permanent residents. The number of students commencing health professional programmes increased over time, while the number and proportion who were international students peaked between 2004 and 2013 (Table 2).

Ethnic group

Analyses of commencing students by ethnic group, and all subsequent analyses presented in this article, are restricted to students who were Aotearoa citizens or permanent residents.

Ethnic diversity increased within health professional programmes (Table 3, Figure 1) and across the wider university (Figure 1) over time. While the proportion of commencing students who were Māori or Pacific increased overall, parity with other ethnic groups was not approached in most instances. The exception to this was in the Bachelor of Medicine and Bachelor of Surgery (Figure 1) where, by the 2019–2023 period, Māori students comprised 20.1% of all incoming students and had commenced at a rate (per 100,000 estimated resident population aged 18–29) comparable with that of Asian and European students (see supplementary material, p. 118, 128). Overall, students within the Asian ethnic grouping comprised a greater proportion of students commencing health professional programmes than other courses within the wider university (Figure 1), although the ethnic composition of individual professional programmes varied widely (supplementary material).

Gender

Students who identified as female predominated within health professional programmes across all years, increasing from 62.3% of commencing students during the 1994–2003 period to 65.0% during 2014–2023 (Table 3). Female students also predominated within the wider university (supplementary material, p. 32). Although a small number of programmes (Bachelor of Dental Technology, Bachelor of Dental Surgery, and Bachelor of Medicine and Bachelor of Surgery) had a roughly equal gender balance or male majority towards the beginning of the study period, female students were a majority across all health professional programmes by the mid-2010s (supplementary material, p. 50). Very few students who identified as another gender were recorded (Table 3), reflecting limitations to data collection processes.

Socio-economic measures

Incoming cohorts were highly skewed towards students from more socio-economically privileged backgrounds. Students from schools in the lowest socio-economic quintile (least socio-economically advantaged) were nearly absent from health professional programme admissions, comprising approximately 2% of students entering those programmes across time (Table 3). A strong predominance of students from higher quintile schools was also observed consistently within the wider university and individual health professional programmes (Figure 2). A similar pattern, although somewhat less pronounced, was observed by NZDep quintile (Table 3).

Urban/rural classification

There was a strong predominance of urban backgrounds among incoming students (Table 3, Figure 3), reflecting the geographic distribution of the underlying populations. The commencement rate (per 100,000 estimated resident population aged 18–29 years) for health professional programmes combined was broadly similar for students with home addresses within the U1 (most urban), U2 and R1 GCH categories throughout the study period (supplementary material, p. 59). While lower initially, the commencement rate for the R2 category increased to meet that of the aforementioned categories over time. The commencement rate for health professional programme students with home addresses within the R3 category remained consistently lower throughout the study period.

Admission category

Approximately two-thirds of admissions since 2017 across all undergraduate health professional programmes within the Division of Health Sciences were for students who had applied via the health sciences first year admission category (Table 4), with the remainder being a mixture of graduate students and students admitted via other pathways, which vary between programmes.

Affirmative action admission subcategories

Admissions under the general category made up the majority of admissions within undergraduate health professional programmes (68.2%) and the Bachelor of Medicine and Bachelor of Surgery (50.9%) since 2017 (Table 4). The largest proportion of admissions via affirmative action pathways was for the Māori subcategory, closely followed by the rural subcategory. A smaller number of admissions occurred via the Indigenous Pacific, socio-economic equity and refugee background subcategories (the latter two subcategories having only commenced in 2020).

Discussion

In this paper and the accompanying supplementary report, we present extensive analyses of the socio-demographic characteristics of students entering all courses, all health professional programmes and 11 individual health professional programmes at Ōtākou Whakaihu Waka between 1994 and 2023. These analyses provide unique insights into who has had the opportunity to study at a leading Aotearoa academic institution over a timeframe spanning a generation.

During this 30-year period, there was a notable increase in the proportion of domestic health professional programme students who were Māori or Pacific, and an increase in enrolments of students from rural backgrounds. At the same time, the socio-economic profile of incoming students remained unchanged, with the students entering essentially all programme categories, across all years, being highly skewed towards those from more socio-economically privileged backgrounds. A steady increase in the proportion of students who were female continued across all years, with nearly two-thirds of all domestic health professional programme students identifying as female by the study end. These findings confirm and extend patterns and trends identified by earlier cross-sectional analyses of health professional students at Ōtākou Whakaihu Waka.^12,13 They also broadly reflect patterns identified by more recent nation-wide, cross-sectional analyses of health professional students undertaken across multiple institutions.²⁸

The findings of this study have important implications. The increase in students who are Māori or Pacific within the Bachelor of Medicine and Bachelor of Surgery is a success, particularly given the earlier barriers that existed for such students.¹¹ For example, by the 2019–2023 period, Māori students comprised 20.1% of all incoming students in the Bachelor of Medicine and Bachelor of Surgery and had commenced at a rate comparable to that of Asian and European students. This increase was a result of stronger recruitment, admissions and student support policies, reinforced by medical school accreditation requirements. Increases were not seen to the same extent across all programmes, however, and in some there was little change. In such instances there is a need for stronger policies to be introduced and evaluated; for example, for outreach and recruitment, bridging/foundation programmes for Māori and Pacific students, equity-focussed admissions policies and student support programmes. Furthermore, while findings for the medical programme are encouraging, Māori and Pacific peoples remain severely under-represented among the current workforce of practicing doctors.²⁹ Given the time scale of medical training pipelines and careers, even if Māori and Pacific students entered Aotearoa medical schools at a much higher rate, inequities in workforce representation would take many years to reduce.

Analyses of socio-economic measures paint a consistently bleak picture of stratified educational opportunity that persisted and, if anything, worsened over time. The near exclusion of students from schools in the lowest socio-economic quintile across essentially all of our analyses is a notable finding that should prompt reflection as to how our education system and society marginalise such students, and fail to enable them to pursue tertiary education at anywhere near the level of their more socio-economically advantaged peers. With the importance of education as a determinant of income,³⁰ such inequities are highly likely to be self-perpetuating. Recently, Ōtākou Whakaihu Waka implemented affirmative action policies and programmes to facilitate entry to health professional programmes for students who attended schools that serve less socio-economically privileged communities.^3,31 However, as only a small number of years have passed since the inception of those programmes it was not possible for our study to provide a meaningful evaluation of their effect.

This study has some important strengths. The large temporal span, and the use of automated processes to replicate detailed analysis across multiple programme categories, has enabled comprehensive statistics across a broad scope of programmes to be produced (supplementary material). This approach will also support the undertaking of updated analyses in future years. There are also limitations. Most notably, while extensive data exploration and checking of statistical output were undertaken, scrutiny of individual numbers and figures for all outputs could not occur to the same level as if bespoke analyses had been developed and undertaken for each individual programme.

Other limitations related to specific variables are outlined in more detail in the accompanying report (supplementary material, p. 2–22). Briefly: Ōtākou Whakaihu Waka currently only stores data on up to three ethnicity classifications for each student (at least six are recommended);¹⁹ the broad level one “Asian” ethnic category has acknowledged shortcomings;³² historic data collection processes meant that gender data were essentially limited to a binary male/female classification across most included years; errors in recording and geocoding home addresses may have translated into small errors in findings for spatially derived measures (NZDep, GCH); use of students’ commencement year to assign NZDep quintiles and school deciles will not have resulted in the most appropriate version being assigned in all cases; use of 2018 GCH classifications across all years means that changes in the rural/urban status of some areas over time will have resulted in some misclassification; and, finally, admission category and subcategory data were only available since 2017.

Nevertheless, our findings provide valuable insights that will support the continuing conduct and development of Ōtākou Whakaihu Waka affirmative action programmes. While efforts to enhance health professional student diversity have had a positive impact, it is clear that the university’s vision of a health workforce that represents Māori and the diverse contexts of our society in Aotearoa³ remains far from being realised. Achieving this vision will require long-term ongoing effort and commitment. The methods that we have developed for evaluating health professional programme student demographics will provide a valuable source of intelligence to support these efforts on an ongoing basis.