The capacity of the public health workforce is a key contributor to the ability of communities to address public health nutrition issues(Reference Baillie, Bjarnholt and Gruber1). Workforce capacity is influenced by a range of determinants including the quality of workforce preparation and continuing professional development, workforce size, organisation and support(Reference Hughes2). With respect to workforce preparation and continuing professional development, competency-based approaches have been widely embraced recently as a process central to the professionalisation of public health and its related disciplinary groups(3–Reference Hughes7), including public health nutrition(Reference Hughes7).
In the European context, workforce development that encourages optimal workforce mobility and collaboration in research and practice requires the development of comparably competent practitioners who are capable of developing and undertaking effective population-based strategies to meet nutrition and public health objectives. This has been the position of workforce development scholars for at least the last decade(Reference Haughton, Story and Keir8, Reference Yngve, Sjostrom and Warm9). Competency standards provide the architecture for workforce development by codifying the knowledge, skills and attitudes necessary to effectively practise public health nutrition(Reference Hughes10). They have a deliberate focus on effective performance in the workplace, ensuring that workforce preparation and continuing professional development not only enhance what practitioners ‘know’, but also that they ‘know how’, can ‘show how’ and ‘do’(Reference Wass, van der Vleuten and Shatzer11).
Competency standards serve a number of workforce development functions, including providing a structure for the design and evaluation of curricula that promote minimum standards, assessing individuals in terms of fitness to practice, directing continuing professional development and assisting job evaluation and design. The utility of competency standards as a tool for workforce development is increasingly being recognised worldwide. Competencies specific to public health nutrition have been of interest to workforce developers in the USA for at least three decades(Reference Haughton, Story and Keir8, Reference Sims12, Reference Story, Neumark-Stzainer and Ireland13). In Australia, advanced-level competencies have been proposed to help delineate the roles and workforce development needs required for this developing workforce(Reference Hughes7) and competencies have been codified as a basis for public health nutritionist registration(14). In Europe, the Nutrition Society in the UK has developed a system for registration of public health nutrition practitioners as a public protection and quality assurance measure(Reference Landman15, 16) and this has included consideration of competency issues(16).
Attempts to assess consensus on the competency needs of public health nutritionists in the early 2000s demonstrated a high level of agreement on essential public health nutrition competencies identified by an international panel of public health nutrition experts from the USA, Europe and Australia(Reference Hughes17). Although panellists were recruited from a limited range of cultural and health system contexts, this finding supported the suggestion that competencies required for effective public health nutrition practice are largely consistent across countries and settings (at least among developed countries) and that workforce development for public health nutrition can be based on a consensus set of competencies that are transferable across jurisdictions and in different contexts(Reference Hughes10, Reference Hughes18). Many of these competencies identified were similar to those of general public health practice(3, 5–Reference Hughes7), but with a consistent provision that the public health nutrition workforce requires additional competency units in nutritional sciences(Reference Hughes18). This earlier scholarship focusing on public health nutrition competencies has promoted a premise that the mix and level of competency required by an individual practitioner will vary in different practice contexts, but that there is a core set of essential competencies required for a practitioner to be assessed as competent(19).
Given this potential utility of competency standards for workforce development efforts with currency in different national contexts, the development of consensus on competency standards in the pan-European context is important. The aim of the present study was to assess and develop consensus among public health nutrition workforce development stakeholders in Europe on essential competencies for the effective practice of public health nutrition, and the level of proficiency required at different levels of practice.
Methods
A modified Delphi study, based on an earlier consensus development study focusing on public health nutrition competencies(Reference Hughes10), was implemented among workforce development stakeholders from across Europe. It included, as shown in Fig. 1, a definition of the problem through gathering of current intelligence, identification and recruitment of panellists, followed by a three-step process for identifying consensus in the panel. The Delphi study’s primary advantages in this context are cost-effectiveness and the ability to harness the anonymous ideas and opinions of a range of experts/stakeholders in different geographic locations. The Delphi method is a frequently used technique to measure and aid forecasting and decision making in a variety of disciplines(Reference Murphy, Black and Lamping20, Reference Rowe and Wright21), including competency requirements in health-care settings(Reference Murphy, Black and Lamping20). Four key features define the Delphi method: anonymity, iteration, controlled feedback and the statistical aggregation of group response(Reference Rowe and Wright21).
Fig. 1 Schematic overview of the Delphi process (adapted from reference (Reference Hughes10); PHN, public health nutrition)
Expert panel recruitment
The Delphi method does not call for expert panels to be representative samples for statistical purposes. Representativeness is assessed on the qualities of the expert panel rather than its numbers(Reference Powell22). There seems to be very little actual empirical evidence on the effect of the number of participants on the reliability or validity of consensus processes(Reference Murphy, Black and Lamping20). Limited guidance therefore exists on the minimum or maximum number of experts on a Delphi panel. It appears to be related to common sense and practical logistics(Reference Keeney, Hassen and McKenna23).
An email contact list for European public health nutrition workforce development stakeholders was developed, primarily through snowball sampling from existing professional networks and a list of contacts developed by internet search of websites of European universities, official websites of public health nutrition-related research projects, conferences and workshops, and through the Nutrition Society website.
The inclusion criterion for individuals invited to become panellists was that they were required to be in positions relevant to influencing and practising workforce development (e.g. employers in public health agencies, academics and senior practitioners). The assumption used when identifying panellists this way was that individuals employed or practising in senior-level positions within health and academic institutions had the necessary experience and insight to be considered experts in this context.
The final contact list comprised 159 identified experts across Europe, who were formally invited via email to join the pan-European expert panel and participate in the Delphi process with an explicit request to either self-nominate or forward invitations to known experts. Panellists who accepted the invitation (i.e. expert self-selection) were forwarded the first Delphi questionnaire as an email attachment for completion and email return. Figure 1 illustrates the process used in this study.
Survey instruments
Each survey round of the Delphi process used survey instruments based on an earlier study that assessed consensus among an international panel regarding the essential competencies required for effective public health nutrition practice. The competency units tested in these survey rounds were originally constructed from competency standards in the public health, health promotion and dietetics fields of practice(Reference Hughes10). The present study varied from the traditional Delphi method in that structured questions informed by this earlier study were used rather than open-ended questions.
Table 1 summarises the focus of inquiry in each survey round.
Table 1 Focus of the Delphi survey round questions used in Rounds 1 to 3
*Essential competencies defined as competencies without which public health nutrition practice effectiveness is limited(Reference Hughes10, 19). Useful competencies defined as competencies that are required occasionally or in specific contexts only.
Ratings of levels of competency required by different workers (round 3)
The fifty-seven public health nutrition competency units tested in Delphi survey rounds 1 and 2 were tested for the level of competency required by three broadly defined workforce groups using the definitions in Table 2.
Table 2 Definitions of worker category and level of competency used in rating competencies in round 3
Data analysis
Responses to each round of Delphi surveys were entered into Microsoft® Office Excel 2003 (Microsoft Corporation, Redmond, WA, USA) for storage. The data were imported to the SPSS for Windows statistical software package version 11·0 (SPSS Inc., Chicago, IL, USA) for determination of frequency of response distributions and χ 2 analysis for differences in distribution between academic v. employer panellists. The level of significance was taken as P < 0·05.
Definition and stability of the consensus
The literature does not provide any agreed standard on how to measure consensus(Reference Rowe and Wright21). Arbitrary and predefined consensus rules were used to judge agreement among the panellists and define when consensus had been developed. The consensus standard was arbitrarily set at ≥66·7 % agreement (i.e. two out of three panellists agreeing) in all three Delphi rounds. The consensus was considered to have reached stability if group ratings on suggested competencies varied ≤10 % between rounds. This approach to define stability of the consensus has been used in previous studies(Reference Hughes10, Reference Duffield24).
Results
Characteristics of the Delphi panel
A total of sixty-two panellists participated in the Delphi study, with fifty-two (84 %) completing all three surveys. Figure 1 provides data on the number of panellists responding in each round. Participating panel members were from twenty countries, seventeen EU countries and three countries which are not EU member states (refer to Acknowledgements: panellists and country for details). Delphi respondents were categorised into two subgroups (employers v. academics) based on their current position. About one-third were categorised as employers (e.g. working in public health institutes, ministries, private companies, etc.) and the rest as academics (e.g. university personnel, employees at research institutes, etc.).
Agreement with basic propositions relating to public health nutrition workforce development
Table 3 presents levels of agreement on a 5-point Likert scale (1 = strong disagreement, 5 = strong agreement) with a range of propositions relating to public health nutrition workforce development and competencies. These results indicate a high degree of panellist homogeneity in terms of agreement with the definition of public health nutrition and propositions relating to public health nutrition workforce development. There was limited difference in response distributions between panellists categorised as academics v. employers, with more academics agreeing with proposition 5 (Table 3).
Table 3 Response distributions for agreement to propositions relating to public health nutrition workforce development and competencies (Delphi round 1, n 60), descending order by percentage of agreement or strong agreement
*5-point agreement/disagreement scale.
†χ 2 = 4·3, P = 0·04, 94·6 % of academics v. 76·2 % of employers. No significant difference in response distribution between academic and employer panellists other than proposition 5 (χ 2, P > 0·05).
Essential (core) public health nutrition competencies
Table 4 presents data on the proportion of panellists rating competency items considered as essential for a public health nutrition practitioner (from Delphi survey rounds 1 and 2) and the most common response (modal response) for predefined levels of competency required for three different workforce tiers (rounds 2 and 3).
Table 4 Consensus on public health nutrition competency units rated as essential (core) and proportion of ratings for level of competency by workforce level (n 52 panellists completing three rounds)
*Bolded competencies for individual practitioner defined as essential by ≥66·7 % at round 2. Essential competency defined as one without which public health nutrition practice effectiveness is limited.
Level of competency by workforce level ratings in bold reached the consensus limit, i.e. ≥66·7 % agreement.
Shaded cells indicate significant difference in response distribution between academics v. employers (χ 2, all P < 0·05), academics more likely to rate as displayed than employers.
Nineteen of the fifty-seven suggested competency units (33·3 %) were rated as essential by more than the consensus standard of ≥66·7 % agreement at the first round. At the second round, twenty-nine competency units or 50·9 % reached the consensus standard and were defined as essential public health nutrition competencies (see Table 4).
Differences in ratings between the two panellist categories (academics v. employers) were limited to only two of the competencies defined as essential. Academics rated two specific suggested public health nutrition competencies as essential more often than employers did. These were Nutrition monitoring and surveillance (χ 2 = 4·7, P = 0·03, 97·1 % of academics v. 78·9 % of employers) and Knowledge of the roles and cultures of other health professions in public health (χ 2 = 4·7, P = 0·03, 88·2 % of academics v. 63·2 % of employers).
The highest agreement among panel members was on competencies clustered within the Analytical, Public health services, Nutrition science and Professional headings. Only one suggested competency unit, Values and participates in peer review under the Professional heading, did not reach the consensus limit.
Response stability between rounds
Only three of the twenty-nine competency units rated as essential at consensus level (≥66·7 %) were response stable after two rounds (i.e. changing ≤10 % between rounds 1 and 2). In all twenty-nine competency units rated essential at the consensus level, this change between rounds was upwards (increasing between rounds 1 and 2), indicating a strengthening of agreement. For all competency units tested, response stability was reached on twenty-one of the total of fifty-seven suggested competency units (36·8 %).
Level of competency required by workforce level
For the specialist work category, all but one of the twenty-nine competency units rated as essential to consensus level were also rated as requiring advanced-level competency. The manager level worker category tended to have a mix of advanced- and intermediate-level competency requirements across competency elements, with a definite clustering of advanced-level requirements in the Communication and Management and leadership competency domains. Front-line workers were rated to need basic competency levels across most competency units tested, with higher-level competencies required in competency units reflecting Professional and Nutrition science competency domains. Ratings of competency level required for different workforce levels showed minor variation between academic and employer category panellists, with only four of fifty-seven units being significantly different in response distribution (χ 2, all P < 0·05, with academics more likely to rate more often than employers).
Discussion
A basic and logical assumption of the sampling method used to develop the Delphi panel was that individuals occupying academic or senior-level professional roles relevant to public health nutrition were ‘experts’ in the context of public health nutrition workforce development. Panellist self-selection after invitation is also likely to engage individuals who consider themselves to be experienced and knowledgeable in this field. The sampling method used in the current study achieved higher initial participation and panellist retention across the three survey rounds than the earlier study by Hughes(Reference Hughes10), on which the current study was based. The good panellist retention across the three survey rounds (n 52, 84 % of panellists) indicates their high degree of interest in the topic. The high degree of agreement among panellists with the definition of public health nutrition and a range of workforce development propositions relevant to public health nutrition competencies suggests that this panel was homogeneous in terms of their vision of public health nutrition and opinions regarding the relevance and importance of competencies as a workforce development issue.
Consulting employers about their competency expectations for public health nutritionists has been proposed as an important part of the intelligence gathering required to inform workforce development(Reference Hughes25, Reference Torheim, Granli and Barikmo26). The assumption is that employers know what competencies are needed of public health nutritionists and/or have adequately reflected on this question. The limited variability in response distributions for competency ratings by academics v. employers throughout the whole study suggests that this categorisation of key stakeholders (academic v. employers) does not isolate differences in experience and opinion, at least in terms of how these are expressed in competency ratings.
As most of the competencies defined as essential (twenty-six of twenty-nine) did not achieve consensus stability over two rounds, further survey rounds would have been required to achieve this aspect of consensus development in the Delphi method. All of the changes between rounds were in favour of increasing agreement, indicating that panellists were focusing their responses consistent with the overall panel. The competency units rated as essential in the current study matched closely the earlier agreement identified in the 2003 Delphi study(Reference Hughes10). Many of the competency units identified as essential in the current study mirror competencies isolated in health promotion(Reference Hyndman27) and public health(28) practice disciplines. This is not surprising given that the original Delphi survey competency units tested were largely derived from the health promotion and public health literature. This is consistent with the almost universal support among panellists for the proposition that there are many cross-cutting competencies between public health and public health nutrition. It also reflects the position suggested in earlier work(Reference Hughes18) that public health nutrition as a field of practice is more closely aligned as a specialisation within public health than as a specialisation within nutrition. This has been reinforced more recently(Reference Cannon29) and is open to debate. Given the nutrition-specific disciplinary focus of public health nutrition, it is not surprising to note the very high proportion of panellists rating competencies within the nutrition science competency domain as essential. This supports earlier opinions expressed about the central importance of competence in nutritional sciences for public health nutritionists(Reference Hughes18).
This attempt to delineate the level of competency required by different workforce categories in the current study recognises the many players in the public health nutrition workforce and the competency development priorities for a specialist workforce. These data should be of value to workforce developers and educators, particularly in the European context where workforce specialisation is limited, because they help prioritise workforce development needs.
The similar views of the two subgroups of panel members support the concept of core competencies recognised by practitioners working at different levels in the public health nutrition field. The high level of agreement on essential competencies among this pan-European panel supports the idea of essential competencies applicable across countries and in different contexts(Reference Hughes10). The strong consensus developed on competencies essential for efficient public health nutrition practice can be used to develop the public health nutrition workforce in Europe, by providing a framework for benchmarking curriculum and assessment among existing specialist education programmes and to inform new programme development. The essential competencies identified also enable practitioners and students to reflect on their own competencies and assess their continuing education needs.
The breadth of the competencies defined as essential in the current study indicates that the focus of workforce development should be on forming efficient work teams or workforce tiers with the required competency mix. The competency priorities identified for specialists also provide direction for the development of a public health nutrition specialist in countries that do not have such roles. This need for specialists (i.e. a designated public health nutritionist) has been a consistent feature of workforce capacity building in other parts of the world such as Australia, the USA and Canada.
The present study provides the data to support the development of a number of workforce development instruments based on consensus and priority in the European context. These competency standards can be used to assist with human resource management (position descriptions, recruitment, etc.). It also builds on the limited workforce development research in the public health nutrition arena, with the potential to contribute to more strategic workforce development in Europe and at a global level.
Lack of information regarding the public health nutrition workforce in Europe is an obvious barrier to workforce development. This Delphi study provides information to aid in the workforce development of public health nutrition practice by gathering expertise and intelligence from a large group of geographically dispersed public health nutrition experts. The results show agreement about essential competencies for the public health nutrition workforce, especially for elements within the nutrition science as well as the professional, analytical and public health service competency domains. The diverse background of panellists gives added value to the statements reaching the consensus limit, further supporting the concept of core competencies and workforce functions transferable between countries.
Acknowledgements
This Delphi study is a part of a pan-European project, JobNut, which was financed by the Leonardo Community Vocational Training Programme (agreement number 2003320). R.H and A.Y. confirm that they had no conflict of interests relating to this research other than their roles respectively as Deputy Editor and Editor-in-Chief of the journal, and had no role in terms of peer review or approval. S.J. and I.T had no conflicts of interest. S.J. participated in the design of the Delphi study and implemented the study as a research student in nutrition. She had a principal role in Delphi data analysis and reporting. R.H. conceptualised the study and had a principal role in methods design and data analysis. I.T. participated in the survey design and data analysis. A.Y. contributed to conceptualising the study and was the project’s manager. All authors contributed to the drafting and final editing of this manuscript. The generous contribution and input from the Delphi panel members is deeply appreciated. Obviously, the current study would not have been possible without the efforts of the stakeholders, giving their time and sharing their expertise. The Expert Panel comprised the following individuals: Katica Antonic Degac (Croatia); Katherine O’Doherty Jensen, Lotte Holm, Sjurdur Olsen (Denmark); Arja Erkkilä, Irja Haapala, Jaakko Kaprio, Pekka Puska (Finland); Ambroise Martin (France); Helmut Oberritter (Germany); Theodora Psaltopoulou, Éva Martos (Greece); Andrea Lugasi (Hungary); Elva Gísladóttir, Ingibjörg Gunnarsdóttir, Laufey Steingrímsdóttir (Iceland); Cecily Kelleher, Corina Glennon Slattery, Eileen Gibney, Geraldine Quinn, John Kearney, Linda Hogan, Nick Kennedy (Ireland); Domenico Palli (Italy); Albertas Barzda, Algis Abaravicius (Lithuania); Yvette Azzopardi (Malta), Alma van der Greft, Annemiek van Bolhuis, Joop van Raaij (Netherlands); Arne Oshaug, Knut-Inge Klepp, Øydis Ueland (Norway); Lucjan Szponar (Poland); Maria Daniel Vaz de Almeida (Portugal); Cirila Hlastan-Ribič, Jozica Maucec-Zakotnik (Slovenia); Carmen Pérez-Rodrigo, Emilio Martínez de Victoria Muñoz, Gregorio Varela Moreiras, Lluís Serra Majem, Victòria Arija (Spain); Agneta Yngve, Anki Sundin, Annica Sohlström, Gudrun Ahlin, Gunnar Johansson (Sweden); Anne Heughen, Ailsa Welch, Alison Nelson, Barrie Margetts, Claire Schofield, Diane Talbot, Elizabeth Lund, Jackie Landman, Jessica Swann, Kim Procter, Rachel Thompson, Sheela Reddy, Sinead McElhone, Susan Jebb, Wendy Wills (United Kingdom).
