To explore Finnish experts' perceptions of the forms of digital healthcare that are anticipated to be the most utilised in healthcare in the medium-term future (year 2035) and anticipated healthcare workforce impacts those forms will have.
A total of 17 experts representing relevant interest groups participated in a biphasic online Delphi study. The results for each round were analysed using descriptive statistical methods and inductive content analysis.
The forms of digital healthcare that the experts perceived as most likely to be utilised were those enabling patient participation, efficient organisation of services and automated data collection and analysis. The main impacts on the healthcare workforce were seen as being the redirection of workforce needs within the healthcare sector and need for new skills and new professions. The decrease in the need for a healthcare workforce was seen as less likely. The impacts were perceived as being constructed through three means: impacts within healthcare organisations, impacts on healthcare professions and impacts via patients.
The results are not necessarily transferable to other contexts because the experts anticipated local futures. Patients' views were also excluded from the study.
Healthcare organisations function in complex systems where drivers, such as regional demographics, legislation and financial constraints, dictate how digital healthcare is utilised. Anticipating the workforce effects of digital healthcare utilisation has received limited attention; the study adds to this discussion.
Introduction
Healthcare systems worldwide are searching for ways to adapt to drivers, such as the need for sustainable, equal healthcare systems, demographic changes and emerging technologies (Braithwaite et al., 2018). One of the key problems has been the availability of a healthcare workforce, which can be broken down into retention problems, difficulties in attracting healthcare professionals, skills mismatches, and inefficient organisation of work, to name a few main issues (WHO, 2022). The utilisation of digitalisation and technology in healthcare has been seen as a potential way of responding to those challenges by providing effective, timely healthcare services through increasing automatisation and patient participation and improving the flow of information and efficacy of time usage (European Commission, 2018; Lapão et al., 2019; Topol, 2019). Because of the relative freshness of the phenomena, there are no established terms to describe it but for example, the term “eHealth” has been proposed (Shaw et al., 2017). The term “digital healthcare” has also been suggested to describe the forms of healthcare that operate within the framework created by technology (Topol, 2019; Vähäkainu and Neittaanmäki, 2018; Zarif, 2022) and will be used in the present article to describe the phenomenon.
Impacts on the healthcare workforce are not an explicit part of healthcare technology assessments (Fattore et al., 2011; Haverinen et al., 2019), despite the hopes of reduced workforce needs placed on the increased utilisation of digital healthcare. The utilisation of digital healthcare is guided by regional and national strategies, national and international legislation, financial constraints and the technological skills of both staff and patients. The position of digital healthcare as a part of the international, national and regional healthcare systems—here with multiple factors influencing its introduction and utilisation—supports viewing healthcare as a part of a complex system (Braithwaite et al., 2018; Lipsitz, 2012; Rouse, 2008). In a complex system, understanding the functioning of individual parts is not sufficient for comprehending the whole system because it is more than the sum of its parts (Puustinen and Jalonen, 2020). It has been suggested that the practice of anticipation is useful when there is a desire to shape the future of complex systems (Pernaa, 2020). Here, anticipation can be described as viewing the future as building on the actions and decisions taken in the present (Miller et al., 2018, p. 53) and as a way in which the future is expressed in the present (Miller, 2018). As the images of the future guide decision-making processes (Linturi and Kuusi, 2022), it is essential to be aware of what those images are. This is what the present study aims to offer because this can provide an additional perspective for digital healthcare utilisation against the backdrop of the decreasing availability of the healthcare workforce.
Digital healthcare can be used to monitor health, enable communications and collect, manage and utilise health data (Shaw et al., 2017), and it has been studied from various perspectives, such as ethics and acceptability (Busse et al., 2021; Zarif, 2022), efficacy (Granja et al., 2018), economical aspects (Tenhunen et al., 2018), equal access (Kaihlanen et al., 2022; Kwiatkowska and Skórzewska-Amberg, 2019) and future prospects (Immonen et al., 2019; Topol, 2019). The impacts of digital healthcare on the workforce skill needs (Bollinger et al., 2013; Radnia, 2018) and the workforce itself have also been discussed. The interview study by Lapão (2016) indicated that increasing the utilisation of digital healthcare will have two main implications: it will reduce the need for healthcare staff by increasing the efficacy, but it will also increase workforce needs as new forms of healthcare services are created. Digital healthcare has also been seen as diverting both service and workforce needs into the primary healthcare sector (Bronsoler et al., 2020). In earlier publications, digital healthcare has been thought as changing the way healthcare staff do their work, rather than changing the actual workforce needs (Meskó et al., 2018).
Previous works that have explored the possible, probable and desirable futures of technology utilisation in healthcare systems have identified the development of artificial intelligence (AI), the internet of Things (IoT) and robotics, improvements in sensor technologies, digital architecture and information management as essential aspects (Jauhiainen et al., 2017; Topol, 2019). However, because technological development in the healthcare sector often follows the technological development of other industries and is utilised in various ways within different parts of the healthcare sector, it is challenging to navigate this complexity to anticipate which forms of digital healthcare will be utilised in the future and what the implications for the healthcare workforce needs will be.
The purpose of the present study was to explore Finnish experts' perceptions of the forms of digital healthcare that are anticipated to be the most utilised in the medium-term future (year 2035) and the likely healthcare workforce impacts those forms of digital healthcare will have. This was done by utilising the Argumentative Delphi method. In an anonymous, biphasic Delphi process, the panel consisting of experts representing healthcare organisations (HCO), healthcare technology organisations (HTO) and healthcare education institutions (HEI) evaluated the likelihood and desirability of different scenarios. The Delphi method can be useful when the studied phenomenon has various possible futures and no specific interest group has certain information about its future developments or even a complete overview of the phenomenon (Linturi and Kuusi, 2022) but there is a need to create preferences between several possible futures (Keeney et al., 2001). As such, it is suitable for studying complex issues in complex systems (Miller et al., 2018, p. 59).
Methods
Study design and setting
The study was conducted utilising the Delphi method, which has been described as a way of having a controlled debate about forecasting future developments when future development is not known and cannot be calculated statistically (Gordon, 2009). The Argumentative Delphi method was seen as an appropriate variation of the Delphi for the purposes of this study, as one aim was to explore the potential differences between the anticipated workforce impacts of digital healthcare amongst the different stakeholders. The study setting was a Finnish province known as a healthcare technology hub. In Finland, both the healthcare sector and society in general are seen as especially prepared to embrace new forms of digital healthcare, partly because of the swift transition of services created by the Covid-19 pandemic, vast healthcare databanks and high-quality education system paired with good digital skills at the population level (Hendolin and Hämäläinen, 2022; Finnish Government, 2020). These qualities made the particular Finnish province a suitable setting to study the phenomena of digital healthcare.
The selection and recruitment of the expert panel
For selecting and recruiting the experts, four main interest groups were identified through stakeholder analysis in co-operation with the municipal authorities: HCO, HTO, HEI and patients. After consideration, the view of the patient-expert was excluded because the expertise of the patient could not be easily identified, a problem shared by health technology assessment initiatives in general (Wortley et al., 2016). For defining expertise, the definition by McPherson et al. (2018) was adopted, in which expertise can be understood as having special knowledge derived from training or experience. Six organisations representing relevant interest groups (HCO, HTO and HEI) within the province were chosen. The recruitment of experts was done by snowballing within those organisations to ensure the coverage of the panel (Stubin et al., 2020). For this, key informants were identified based on their expertise and position within the organisation in relation to the topic and the snowballing was done through them. Participating experts were asked to identify their expertise from five predefined areas: technological development, service provision, ethics, service planning and development and legal expertise. The completed expert matrix (Table 1) shows that the coverage of the panel was sufficient for the purposes of the study.
Matrix displaying self-assessed expertise of the participating experts
| Expertise/Interest group | Technological development | Providing services | Ethics | Service planning and development | Legal expertise |
|---|---|---|---|---|---|
| HCO 2 | x | ||||
| HCO 1 | x | ||||
| HCO 2 | x | x | x | ||
| HCO 2 | x | x | x | x | x |
| HCO 2 | x | ||||
| HCO 2 | x | x | x | x | |
| HCO 1 | x | x | |||
| HCO 2 | x | x | x | ||
| HCO 1 | x | ||||
| HTO/HCO 2 | x | x | x | x | x |
| HCO/HEI 1 | x | ||||
| HEI 1 | x | ||||
| HEI 2 | x | x | x | x | x |
| HTO 2 | x | ||||
| HTO 2 | x | ||||
| HTO 2 | x | ||||
| HTO 2 | x | x | x |
Note(s): Abbreviations: HCO = Healthcare organisation, HTO = Healthcare technology organisation, HEI = Healthcare education institution, 1 = participated in first round only, 2 = participated in both rounds
Source(s): Authors' own work
The experts were contacted by email and received two reminders during each round, here according to a separate communications plan. Initially, 19 experts agreed to participate, but 17 (13 females and 4 males) took part in the first round and 12 (9 females and 3 males) in the second. In a qualitative Delphi study, the size of the panel is not seen as an essential criterion (Stubin et al., 2020), and the saturation of comments can be taken as a sign of a sufficient number of experts (Tapio et al., 2011). In general, a panel of 12–15 experts can be viewed as sufficient (Salkind, 2013). The experts' position in their respective organisations varied between senior management and operative level.
Ethical considerations
The study was conducted according to the principles of the Helsinki Declaration (The World Medical Association, 2013). Research permits were gained from the participating organisations. The experts received an information letter stating that participation was voluntary and could be withdrawn at any time; they then gave informed consent by registering for the study. According to the principles of the Delphi method (Gordon, 2009), the experts remained anonymous during the study, and scientific rigour was followed to protect their anonymity during data analysis.
Data collection
The eDelphi application (www.edelphi.org) was used in data collection. The application also served as a way of managing data during the study by creating automated statistics of the responses and displaying both the statistics and anonymous open comments to the experts during the rounds. The experts were therefore able to both read the comments of others and to respond to them or to adjust their own respond if they wished. The first Delphi round was open online for two weeks, and the second round opened a week later and was also open for two weeks. The opening time of each round was extended by two days because of requests from experts. The first Delphi round focused on experts' perceptions of digital healthcare utilisation in 2035. The experts were first asked to evaluate the likelihood and preferability of three scenarios modelling the development of digital healthcare by the year 2035 using a 7-point Likert scale ranging from extremely unlikely/undesirable (−−−) to extremely likely/desirable (+++). This served as way to introduce the experts to the general topic before the next part, in which they were asked to anticipate the percentual increase or decrease (on a scale of +100% to −100%) of the utilisation of 13 specific forms of digital healthcare in three time points (2025, 2030 and 2035) compared with their understanding of the utilisation of the specific form of digital healthcare in the year 2020.
The forms of digital healthcare under evaluation in the first round were selected using previous research (Jauhiainen et al., 2017; Radnia, 2018; Topol, 2019). Additional data highlighting the current utilisation of the form of digital healthcare were attached to each question. Another option for selecting the forms of digital healthcare to be included would have been to utilise statistical projections on the current forms of digital healthcare and extrapolate their known workforce impacts. Two main problems were identified with this option. First, despite the sizeable Finnish healthcare data pools, there were no comprehensive, reliable and up-to-date recordings of the current levels of digital healthcare utilisation that could have been used to set the baseline. Second, including only those forms of digital healthcare currently utilised would have excluded those forms of digital healthcare not currently in mainstream use. Each individual question received 10–15 answers, and 103 open comments were given in round one.
The second Delphi round consisted of six scenarios set in 2035, in which the forms of digital healthcare that surpassed the limit of inclusion were incorporated. Originally, the limit of inclusion was set to ≥50%, but because only one form of digital technology surpassed it, the limit was reduced to ≥40%. The experts were asked to evaluate the likelihood of each scenario on a 5-point Likert scale ranging from very unlikely (−−) to very likely (++), here from the perspective of three workforce outcomes:
Whether the scenario will reduce the need for healthcare workforce
Whether the scenario will redirect the healthcare workforce needs within the workforce
Whether the scenario will create the need for new professions within the healthcare workforce
These perspectives were formulated based on earlier studies (Bollinger et al., 2013; Bronsoler et al., 2020; Lapão, 2016; Lapão et al., 2019; Lapão and Dussault, 2017; WHO, 2016). A total of 49 comments were provided during the second round.
Data analysis
The data were exported into Microsoft Excel and analysed quantitatively and qualitatively. The means and standard deviations were calculated for the data from the first round to select the forms of digital healthcare that can form the scenarios for round two. Inductive content analysis (Kyngäs, 2019) was utilised in analysing the comments from round one for identifying any essential forms of digital healthcare that might have not been included and those factors that were seen as impacting the utilisation of digital healthcare. In round two inductive content analysis was used for identifying in which ways the anticipated impacts on workforce needs were viewed as constructing. The unit of analysis was defined as a sentence, which was then turned into open codes and grouped to form subcategories, categories and main categories (Kyngäs, 2019).
Findings
Round one
As presented in Figure 1, the experts perceived digital healthcare in general as being a likely and desirable part of the healthcare systems in 2035, with structural barriers being the most potential barriers to this development.
Digital healthcare is an integral part of all forms of healthcare in 2035 on a scale of extremely unlikely/undesirable (−−−) to extremely likely/desirable (+++)
Digital healthcare is an integral part of all forms of healthcare in 2035 on a scale of extremely unlikely/undesirable (−−−) to extremely likely/desirable (+++)
The experts' estimations of the percentual increase or decrease of the utilisation of 13 specific forms of digital healthcare in the 2035 time point are presented in Table 2 in increasing order based on the experts' view of their utilisation. The forms of digital healthcare that were seen as most likely to be utilised were the ones that enabled patient participation, the efficient organisation of services and automated data collection and analysis.
Time point questions from the first Delphi round
| Utilisation of digital healthcare compared with 2020 | Number of answers | 2035 mean* | Standard deviation |
|---|---|---|---|
| 1. Utilising mobile games in self-care guidance | 12 | +51.2% | 18.9 |
| 2. The portion of contacts to healthcare that are handled without physical visit to a healthcare facility | 13 | +49.6% | 25.2 |
| 3. Utilising AI in personal health data analysis | 11 | +46.8% | 25.3 |
| 4. Utilising at home sensors and wearables for diagnostics and remote monitoring | 13 | +46.3% | 16.5 |
| 5. Utilising AI for automated image interpretation | 12 | +45.4% | 17.5 |
| 6. The portion of visits related to chronic conditions being replaced by telehealth contacts and remote monitoring | 13 | +44.2% | 19.1 |
| 7. Utilising AI in healthcare systems data analysis | 12 | +44% | 26.1 |
| 8. eVisits utilising VR and AR | 12 | +42.7% | 25.8 |
| 9. The portion of homecare visits replaced by eVisits and remote monitoring | 12 | +34.6% | 17.9 |
| 10. Utilising robotics in homecare | 12 | +32.9% | 14.3 |
| 11. Utilising AI-based therapy apps | 10 | +30.8% | 15.7 |
| 12. Utilising VR and AR for rehabilitation at home | 12 | +24.2% | 18.3 |
| 13. Utilising genome reading for tailored medicine | 12 | +23.4% | 16.6 |
Note(s): *The range for answers was −100%–100%
Abbreviations: AI = Artificial intelligence, VR = Virtual reality, AR = Augmented reality
Source(s): Authors' own work
During the inductive content analysis of round one, 48 open codes derived from the comments were formed into 21 subcategories, eight categories and three main categories (Table 3) of the factors impacting the forms of digital healthcare that were anticipated to be the most utilised regionally in 2035. The three main categories were organisations, humans and technology. One new form of digital healthcare emerged from the comments and was included in round two: digital care paths.
Difficulties in transferring information across organisational limits will continue to be a problem. (Expert 3, healthcare organisation)
Experts' perceptions of factors impacting the forms of digital healthcare anticipated to be the most utilised regionally in 2035
| Subcategories (n = 21) | Categories (n = 8) | Main categories (n = 3) |
|---|---|---|
| Secondary-level healthcare Primary-level healthcare Private healthcare | Levels of healthcare | Organisations |
| Crossing organisational limits Fading organisational limits | Organisational limits | |
| Laws Information Equality | Structural barriers | |
| Patient-centred care Significance of human contact Technological skills | Patients | Humans |
| Professionals and technology working together The presence of a professional | Professionals | |
| Resistance Unpredictable events Outdated technology Unfinished systems | Problems with technology utilisation | Technology |
| Cost Support services Meeting needs | Accessibility | |
| Affordable equipment Equipment for loan/lease | Availability |
Source(s): Authors' own work
Within organisations, different levels of healthcare and the limits between them and the surrounding structures were seen as the key factors for digital healthcare utilisation. The movement of information and coordination of care between primary- and secondary-level healthcare was also seen as an important factor, while the division between the public and private healthcare sectors was seen as a potential threat. Structural factors, such as information security and the requirement for equal access to services, were perceived as impacting digital healthcare utilisation. Some experts commented that digital healthcare may result in unequal access to healthcare services to some demographic groups, such as the elderly and immigrants.
Technology gets outdated fast and might get too expensive. (Expert 9, healthcare organisation)
The availability and accessibility of technology were seen as important requirements for the integration of digital healthcare. Expenses and the fast rate of technological development were identified as potential problems for technology utilisation for organisations making large investments in technology and digital systems. The absence of support services and introduction of unfinished systems were also perceived as potential obstacles, as well as unexpected events such as changes in geopolitics and counter-reactions to “digital hype”.
It is most important to introduce functions that benefit the patient. (Expert 4, healthcare technology organisation)
The impact of the human factor was perceived as being displayed through both the professionals and patients. Technology and digitalisation were seen as working best in unison with the professional to ensure the best available treatment for the patient. The experts expressed concerns for situations where care processes are centred around digitalisation and technology rather than the patient. Some experts estimated that both patients and professionals will possess sufficient technological skills by 2035, while others perceived insufficient skills as a potential obstacle.
Round two
The scenarios from the second round and likelihoods of their perceived healthcare workforce impacts are presented in Figure 2. The potential of digital healthcare to redirect healthcare workforce needs was seen as mostly likely, and this redirection was seen as happening from one profession to another, but also from the secondary level of healthcare to the primary level of healthcare. The impact that was seen as the most likely was the need for new healthcare professions, especially in developing and maintaining technologies and digital services, as well as data analysts and technical support personnel. The potential of digital healthcare to reduce the need for a healthcare workforce was viewed mostly as unlikely or very unlikely; in contrast, some experts viewed a potential need to increase healthcare workforce because of rising service needs. Some potential to reduce workforce needs was identified in those forms of digital healthcare aimed at assisting in diagnostics or organising and redirecting resources.
The scenarios of the second Delphi round and likelihoods of their perceived workforce impact on a scale of very unlikely (−−) to very likely (++)
The scenarios of the second Delphi round and likelihoods of their perceived workforce impact on a scale of very unlikely (−−) to very likely (++)
In the inductive content analysis process, the replies to each scenario were first analysed individually. During the process, the same subcategories were repeated and towards the final scenario, saturation of the data became apparent because no new categories emerged and the final analysis was performed by combining the comments from all scenarios. There were 76 open codes divided into 31 subcategories, eight categories and three main categories (Table 4). The three main categories through which the workforce impacts of digital healthcare were viewed as occurring were those for healthcare organisations, the impacts on healthcare professions and impacts via patients.
There will be an increased need for primary-level healthcare workers, as patients will be increasingly directed to the primary-level clinics … (Expert 8, healthcare organisation)
Experts' perceptions of factors influencing the impact of digital healthcare on healthcare workforce needs
| Subcategories (n = 31) | Categories (n = 8) | Main categories (n = 3) |
|---|---|---|
| Beginning of the care process Enhancing assessments | Care process | Impacts within healthcare organisations |
| Resources for following alerts Resources for tracking information Smarter resourcing Flexible resourcing Organising resources Enabling needs-based use of resources | Resource management | |
| Decreasing needs in time Decreases need in specialty areas Increased need for services Need for primary-level healthcare workers Needs between professions will shift Needs will change within a specialty | Changing needs | |
| Professions that will understand technology and content New professions for maintaining systems Digital care path developers Digi mentors Digital solution developers | New professions | Impacts on healthcare professions |
| Multiskilled people Data analysis skills Technical skills and ability to notice mistakes Using information for management | New skills | |
| Access to care Access to healthcare professionals Speed of access | Access | Impacts via patients |
| Awareness of health status Awareness of methods Increasing self-care because of awareness | Awareness | |
| Capability to use digital healthcare Capability for individual access to information | Capabilities |
Source(s): Authors' own work
The impacts for healthcare organisations and subsequent impacts on healthcare workforce needs were divided into three subcategories: care processes, resource management and changing needs. The care processes were seen as being impacted by digital healthcare in that they could be more individual and efficient, which was seen as redirecting workforce needs by, for example, reducing the time spent assessing care needs so that it could be spent on delivering care. Digital healthcare was seen as having the potential to enable smarter, flexible resourcing and creating priorities, which could impact healthcare workforce needs in specific units and parts of the care process. Thus, this was seen as changing the need for workforce within different levels of healthcare, focusing the need more firmly on primary-level healthcare. It was also anticipated to enable the division of work differently, such as the transmission of tasks from radiologists to professionals preparing radiological procedures. Digital healthcare was perceived as potentially increasing the demand for services and creating the need for new kinds of services.
… we will need people with multicompetences, people that know how to implement technical solutions to be part of the services, that know how to develop digital services … (Expert 6, healthcare organisation)
The impacts on healthcare professions were divided into new professions and new skills. Current healthcare professions were not perceived as having the required skillsets for the future. The experts estimated that healthcare professionals' skill requirements would be expanded as the need for in-depth technological understanding increased, but also that completely new professions focused on implementing, improving and developing digital health services would be required by the healthcare organisations.
… patients will be more aware of their health status; this will probably reduce the number of unnecessary controls and visits to healthcare “just to make sure” … (Expert 7, healthcare technology organisation)
Impacts via patients included the ways in which digital healthcare's impacts on patients were seen as influencing healthcare workforce needs. Digital healthcare was seen as impacting access to healthcare services by improving availability; on the other hand, it was deemed important that patients should also have access to in-clinic visits and could choose the method of contact. Some experts expressed worries that technology would create distance in the relationship between the patient and healthcare professional, even producing negative health outcomes for the patient, which could lead to an increase in the need for healthcare services and, therefore, workforce needs. Digital healthcare was also viewed as potentially increasing the awareness of health-related issues that could empower and motivate patients to improve self-care, thus reducing the need for healthcare services. On the other hand, the increasing amount of health data produced by patients was seen as potentially increasing the workload of the healthcare systems. An essential aspect was also the capability of the patients to utilise digital healthcare; individual capabilities needed to be recognised and sufficient support services should be available to ensure equal access.
Discussion
The present study contributes to the discussion of the impacts of digital healthcare on the healthcare workforce, producing new knowledge about the forms of digital healthcare experts anticipate being utilised in 2035 and anticipated impacts on the healthcare workforce. The value of utilising anticipatory practices in healthcare service planning is that they can bring additional perspectives into situations where the cost of real-life tests can be high (Miller et al., 2018, p. 54). In addition, being aware of the underlying anticipatory assumptions can empower people to understand how and why the images of the future influence their decisions in the present and to ask new questions concerning the future (Ehresmann et al., 2018, pp. 66–67). The need for healthcare workers is predicted to rise globally because of demographic changes, increases in noncommunicable diseases and an ageing workforce (WHO, 2018), so it is critical to assess the impacts of workforce needs when designing strategies for digital healthcare integration.
According to our results, the integration of digital healthcare into all areas of healthcare by 2035 is a likely and desirable development. The forms of digital healthcare that are perceived as the most likely to be utilised were the ones enabling patient participation, the efficient organisation of services and automated data collection and analysis. This was consistent with earlier studies (Jauhiainen et al., 2017; Topol, 2019) but with some exceptions; the utilisation of technologies such as robotics, virtual reality and augmented reality and applications of genome technology was perceived as less likely, possibly because of the regional drivers for development that the experts were taking into consideration. Previous studies (Habran et al., 2018) have discussed the impact of the driver of development on the forms of digital healthcare. When the driver is technological advancement, development often overlooks existing, similar forms of digital healthcare while ignoring the cost effectiveness of further development of current technologies vs developing new technologies. On the other hand, when the focus has been on the needs of healthcare professionals, the development has leaned on utilising current technologies in new ways or with different patient groups, hence solving specific health problems (Habran et al., 2018). Other studies have also recorded this difference between professionals with different types of expertise (Fruehwirt and Duckworth, 2021).
In the present study, the experts perceived both organisational needs and technological possibilities as the main drivers for the implementation of digital healthcare as problematic, similar to previous findings (Fruehwirt and Duckworth, 2021). The human drivers, especially the patient-centred view and empowering the patients, were seen as being important. The agency of the patient has been recognised as a factor influencing, for example, future hospital management needs (Pihlainen et al., 2019), and the results of the present study have indicated that the patient view is also essential when assessing the drivers of digital healthcare utilisation.
The results suggest that digital healthcare utilisation might not have the desired impacts on healthcare workforce needs because the potential of digital healthcare to reduce the need for healthcare workforce has been perceived as unlikely. Earlier studies have also found digital healthcare as being perceived as changing the way work in the healthcare sector is done (Bronsoler et al., 2020; Meskó et al., 2018). In our study, digital healthcare was perceived to redirect healthcare workforce needs from one profession to another, but also from the secondary to primary level of healthcare. The most likely impact of digital healthcare on the healthcare workforce was perceived as need for new healthcare professions. This has been supported by earlier studies (Lapão, 2016), bringing in the question of how the healthcare sector, which is already struggling to maintain a sufficient workforce, will manage to attract those professionals needed in digital and technological services.
The results suggest that the technical and digital skills of the healthcare workforce are quintessential when following the anticipated integration of digital healthcare. Previous studies have suggested that the digital health training initiatives of healthcare workers should focus on competencies relevant to a particular healthcare setting (Nazeha et al., 2020), such as on nursing documentation and the principles of nursing informatics for nurses (Egbert et al., 2019). Contrary to this, the results support the view that it is important that healthcare workers also have a core set of transferrable digital and technical skills enabling them to work in unison with the professionals developing technological and digital aids and, most importantly, the patients (Li et al., 2019; Risling, 2017). Healthcare workers' digital competencies have also been found to be linked to psychological and organisational factors (Bronsoler et al., 2020; Konttila et al., 2019), highlighting the complexity of developing suitable skillsets for future healthcare workers.
Study limitations
One of the main challenges of the Delphi method is the conflicting expectations for evaluating the reliability of the method and results (Hasson and Keeney, 2011; Tapio et al., 2011). The essential qualities for successful application specifically for healthcare research have been suggested as the identification of the problem area of research, concise definition of expertise and selection of the panel, data management and controlled feedback, hence maintaining suspense during the iterative Delphi rounds, analysis, closing criteria and stability of the results (McPherson et al., 2018; Nasa et al., 2021; Trevelyan and Robinson, 2015). The present study has aimed to address each of these areas, but as in any qualitative study, it is possible that the qualities of the researcher have guided the interpretation of the data and that further study would be required to ensure the transferability of the results. It is also notable that the response rate of the second round fell below 70%. However, as Tapio et al. (2011) have argued, the main goal of a Delphi study is not to produce statistically representative data but to solicitate the opinions of a selected group of experts because any results concerning the future—and the future itself—cannot be known.
The current study has several limitations, most of which deal with the broad view on the issue adopted, which has resulted in the superficial treatment of certain important issues that would have warranted more attention to detail. These include, but are not limited to, the specific forms of digital healthcare and the details of their implementation. The Topol Review (Topol, 2019) identifies and classifies the forms of digital healthcare that are seen as the most essential for the NHS in the United Kingdom by the year 2040 and has used this as a basis for making recommendations for patients, professionals and healthcare systems. Such a comprehensive study was not available for the Finnish healthcare system, and as such, the current study does not necessarily competently reflect the current trends of digital healthcare within Finland. The WHO suggests literature reviews and opinion surveys of key informants as information sources to inform the planning of technological changes in healthcare (WHO, 2022), and the latter is what the present study aims to offer. To be used as background information in policy making, the study would need to be more representative of the healthcare sector. Because this study was limited in its focus to one province of Finland, the topic would also require further study to explore whether these findings are applicable to other regions nationally and internationally.
Conclusion
Healthcare organisations operate as parts of complex systems, and anticipating the impacts of interactions is not simple, nor are the outcomes necessarily straightforward. However, the increasing lack of a healthcare workforce suggests that the impacts of digital healthcare on healthcare workforce needs is an important issue that should be taken into consideration when regional, national and international healthcare strategies are formed. The findings suggest that the technologies that hold the most potential for reducing workforce needs are those focused on enabling patient participation, the efficient organisation of services and automated data collection and analysis. The study also highlighted both the need for new skillsets for healthcare professionals and for new professions that would support both clinical workers and patients in utilising digital healthcare.
The present study confirmed the position of digital healthcare as part of complex systems, and further study is suggested to better identify the future developments of digital healthcare and clarify the key components shaping the workforce impact of digital healthcare. The Delphi method was found to be a relatively inexpensive and fast method of gauging expert opinions, and the use of Delphi as a tool for policy making in healthcare is also a suggested area of further study.
