Healthcare data scientist qualifications, skills, and job focus: a content analysis of job postings

Abstract

Objective

Growth in big data and its potential impact on the healthcare industry have driven the need for more data scientists. In health care, big data can be used to improve care quality, increase efficiency, lower costs, and drive innovation. Given the importance of data scientists to U.S. healthcare organizations, I examine the qualifications and skills these organizations require for data scientist positions and the specific focus of their work.

Materials and Methods

A content analysis of U.S. healthcare data scientist job postings was conducted using an inductive approach to capture and categorize core information about each posting and a deductive approach to evaluate skills required. Profiles were generated for 4 job focus areas.

Results

There is a spectrum of healthcare data scientist positions that varies based on hiring organization type, job level, and job focus area. The focus of these positions ranged from performance improvement to innovation and product development with some positions more broadly defined to address organizational-specific needs. Based on the job posting sample, the primary skills these organizations required were statistics, R, machine learning, storytelling, and Python.

Conclusions

These results may be useful to organizations as they deepen our understanding of the qualifications and skills required for data scientist positions and may aid organizations in identifying skills and knowledge areas that have been overlooked in position postings.

INTRODUCTION

Growth in big data in health care and its potential impact on the industry have driven the need for more healthcare workers and researchers trained as data scientists capable of working with big data. Big data is complex, as it consists of heterogeneous and unstructured datasets, which may include text, images, and video, across multiple areas. Researchers have estimated that unstructured data represent approximately 95% of big data.¹ Analytic techniques such as machine learning and artificial intelligence are used to enhance analysis of both structured and unstructured data.² Critical skills required of data scientists include understanding data heterogeneity and fragmentation as well as data management and conceptualization.³

Big data can be used by healthcare organizations to find innovative solutions as well as to improve care quality and efficiency. Innovation stems from the opportunity to combine traditional data with new data forms, for example, integrating population clinical data with genomics data to improve drug therapies.⁴ To that end, open-access initiatives are underway to increase availability of a variety of data sources for data sharing.⁵ Additionally, big data could be used in innovative ways to achieve greater efficiency, thereby reducing U.S. healthcare expenditures.⁶ There are 5 key areas of savings potential: clinical operations (eg, comparative effectiveness research), research and development (eg, precision medicine), new business models (eg, online platforms), payment (eg, fraud detection), and public health (eg, improved surveillance).⁶ In sum, the work of data scientists holds the potential to have significant impact in health care.

With this in mind, I analyze the content of healthcare data scientist job postings to identify the qualifications and skills required for this work. These research results can be used by healthcare organizations for human resource planning and by those providing education and training to target efforts toward needed skills.

Background

A data scientist is a practitioner who has the knowledge and skills to do data science work. As noted by the National Institute of Standards and Technology Big Data Working Group, data scientists extract knowledge from data to drive action.⁷ Important knowledge areas include statistics (ie, a working knowledge of probability, distributions, hypothesis testing, and multivariate analysis); computer science, which encompasses an understanding of data structures, algorithms, and database systems (eg, Hadoop); and problem formulation (ie, the ability to formulate problems to bring about effective solutions).² Data scientists differ from data engineers in that data scientist core expertise is in math, statistics, and machine learning, whereas data engineer core expertise is in advanced programming and distributed systems.⁸ Machine learning skills, in particular, are becoming mandatory for data scientists for building automated decision systems that provide future predictions. The ability to mine text is also a prerequisite for working with unstructured data, particularly in health care, where much of the clinical data is in a note format.

Data scientists’ defining feature is their ability to go broad (eg, full data analysis cycle) as well as deep for at least 1 aspect of the field such as statistics or big data. This breadth and depth combo is often called “T” formation in which the breadth of skills is represented by the horizontal line and the depth of skills represented by the vertical bar of the T.⁹ Given the nature of the work, data science is an inherently collaborative and creative field, as these scientists generally work within interdisciplinary team environments to find innovative ways to complete projects.⁹ Data communication is also an important skill and includes expertise with visualization tools used to explore raw data as part of the iterative data science process.¹⁰

A number of reports have profiled data scientist roles across industries in an attempt to categorize positions with common focus areas. For example, O’Reilly Strata industry research defined 4 data scientist profession profiles (data businessperson, data creative, data developer, data researcher) and then mapped these profiles to a basic set of technology domains and competencies based on the data scientist practitioners’ self-identification.⁹ In another report, Burning Glass Technologies identified 6 job categories in the data science and analytics landscape (data scientists and advanced analysts, data analysts, data systems developers, analytics managers, functional analysts, and data-driven decision makers) based on similarities in skillsets and functional roles.¹¹ Burning Glass Technologies reported that data scientists require the strongest analytical skills as well as proficiency with specialized tools such as machine learning, Apache Hadoop, and data mining. Moreover, data scientists require generalized skills such as SQL, R, and data analysis.¹¹ In their data science survey, Strata O’Reilly found that Python and Spark were among the tools that contribute most to data scientists’ salary while SQL, Excel, R, and Python were the most commonly used tools.¹²

Demand for data science professionals continues to grow rapidly. Recent cross-industry estimates project demand for data scientists to grow by 28% by 2020.¹¹ Employers are struggling to meet demand for data scientists.¹¹ This skill shortage is compounded by the hybrid nature of data scientist positions; that is, needing a mix of analytic skills and domain-specific expertise, which is difficult to develop in 1 individual.¹¹ This difficulty in finding qualified data scientist candidates is leading organizations to seek creative ways to develop and grow workforce talent in-house.

Healthcare organizations need to formulate strategies to use big data analytics more effectively to achieve healthcare transformation.^13–15 Training staff to use big data analytics is one recommended strategy for doing so.^16,17 Having repeated exposure to the data science life cycle (eg, posing a question, collecting data, exploring the data, developing models, making inferences, and communicating results) helps develop data acumen.¹⁸ Researchers have identified many healthcare big data use cases, including analyzing care patterns and unstructured data, building predictive models, and providing decision support¹⁶; knowledge generation and dissemination, patient engagement, and personalized medicine¹⁴; risk and resource use predictive modeling; and population management,¹⁹ all with potential impact for healthcare delivery.

Research questions and significance

While the work of healthcare data scientists has increased in importance as outlined previously, no research to date has been conducted specifically with regards to healthcare data scientist positions. The purpose of this study was to identify, based on data scientist job postings, the qualifications and skills required for healthcare data scientist positions. This study was guided by the following research questions:

• What are the types of data scientist positions for which U.S. healthcare organizations are hiring and what is the focus of the work?

• What job qualifications and skills are required for data scientist positions in U.S. healthcare organizations?

• How do U.S. healthcare data scientist job qualifications and skills vary by job focus area?

Materials and METHODS

The research questions were addressed by analyzing the content of healthcare data scientist job postings which contain the qualifications and skills that employers require in relation to a specific position. Content analysis is a method for systematically describing documents and written communications. The objective is to condense a volume of text through identifying and describing meaningful categories, analyzing patterns, and thereby achieving new insights.^20–23 Content analysis of job announcements has been used in a variety of professions to evaluate hiring trends^24,25 as well as for research in health care.^16,26 Content analysis consists of the 3 phases of preparation, organization, and reporting.²⁰Figure 1 outlines the steps followed for the content analysis process in this study which were adapted from previous research.^20,21,27,28

Figure 1.

Content Analysis Process.

Preparation phase

The preparation phase includes defining the data collection method, sampling strategy, and unit of analysis.^20,21 The data collection method in this study consisted of capturing active job postings on Indeed.com for U.S. healthcare data scientist positions from February to April 2018. Indeed.com aggregates jobs from organization websites and job boards. The text of each job posting was copied into a separate word-processing document and saved with the job title and organization in the filename.

The sampling strategy was based on a convenience sample of job postings from Indeed.com for healthcare organizations such as health systems, hospitals, insurance companies, vendors, and recruiters. Inclusion criteria for the sample were having “data scientist” in the job posting job title for healthcare organization positions. All active job postings that met the criteria were used in the analysis. Job postings were captured daily and spot checked for position type and organization. Duplicate job postings were eliminated. Job postings were reviewed during the data collection period to assess progress towards data saturation (ie, an indication of optimal sample size).²⁹ After 3 months of capturing job postings, few, if any, new types of positions were identified signaling data saturation had been reached. A total of 198 job postings were identified during the specified time period. The unit of analysis was an individual job posting. The data were reviewed multiple times to better understand the data overall as well as job posting components.

Organization phase

Both inductive and deductive approaches were used in the organization phase. In an inductive content analysis, the concepts are generated from the data and thus adhere to the naturalistic paradigm.^27,30 Deductive content analysis is typically used to evaluate data in a new context or to test categories.²⁰ The inductive approach was used in this study to identify and categorize core job posting information while the deductive approach was used to evaluate skills required for specific job categories.

The inductive content analysis process includes open coding, categorization, and abstraction.²⁰ Open coding was completed by highlighting key parts of the document during the initial review, inserting comments via word-processing software, and identifying meaning units or codes for portions of the text. Category creation is a core feature of qualitative content analysis.³¹ Categories help increase understanding of the research topic and generate knowledge.³² From descriptive headings that were created to describe the content, possible categories were generated and recorded in a spreadsheet. Job postings were reviewed again to abstract basic information such as job title, degree requirements, and years of experience. Iterative reviews were conducted to identify classifications by type of hiring organization, job title, and job focus. Each job focus area was then analyzed in further detail.

Deductive content analysis included developing structured matrices based on skills and coding according to the matrix categories. The skills list was derived from previous data science surveys^9,12 and included both general and specific skills. Skills were coded based on job posting required qualifications and key responsibilities. Coding rules consisted of identifying key terms that matched in the job posting and skills matrix, and then reviewing the related description in the job posting to ensure that the context was appropriate. For example, this job posting text “research, design and prototype robust and scalable models based on machine learning, data mining, and statistical modeling to answer key business problems” was matched to this skill “developing prototype models.” An initial coding was completed for each of the data scientist positions to ensure skills were holistically captured. Only qualifications that fit the skill categories were coded. Recoding of job skills was conducted again approximately 30 days after initial coding and any discrepancies were noted and reanalyzed.

Reporting phase

Descriptive statistics were generated for the overall sample as well as for identified skills. All data for each category were included in the frequency counts representing 100% of the sample. Frequency counts by job levels, hiring organization type, and job focus areas were also generated to allow for more detailed analysis.

Trustworthiness

The qualitative content analysis process must be described in sufficient detail to establish trustworthiness.²⁰ Various aspects of trustworthiness include credibility, confirmability, and dependability.^21,22,31 In this study, trustworthiness was addressed in the following ways. Credibility was demonstrated through data analysis focused on the research questions. Categories were chosen that reflect the research topic and covered the data. Confirmability of findings was demonstrated through tables, a figure and quotes that show a direct link between the results of the study and the data, thus providing an audit trail.³³ Dependability was addressed by collecting the data during a specific time period to ensure consistency and by the use of a code-recode strategy to help address intrarater reliability,^27,34,35 specifically, consistent content categorization using the coding protocol.³⁶

RESULTS

Descriptive statistics

Descriptive statistics for the overall sample based on the inductive analysis are shown in Table 1.

Table 1.

Sample descriptive statistics (N = 198)

	Frequency	Percentage
Hiring Organization Type
Academia	8	4.0%
Biotechnology	4	2.0%
Consulting	12	6.1%
Health System	32	16.2%
Home Health	1	0.5%
Hospital	5	2.5%
Insurance Company	37	18.7%
Pharmacy/Pharma	9	4.5%
Physician Practice	1	0.5%
Recruiter	13	6.6%
Research	9	4.5%
Vendor	67	33.8%
Total	198	100.0%
Degree Required
Advanced Degree	8	4.0%
Bachelor’s and Above	82	41.4%
Master’s and Above	76	38.4%
PhD and MD or Equivalent	18	9.1%
Not Listed	14	7.1%
Total	198	100.0%
Degree Preferred
Advanced Degree	7	3.5%
Master’s and Above	38	19.2%
PhD	32	16.2%
Not Listed	121	61.1%
Total	198	100.0%
Job Level
Data Science Associate	11	5.6%
Data Scientist	125	63.1%
Senior Data Scientist	60	30.3%
Manager/Director	2	1.0%
Total	198	100.0%
Job Focus Area
Innovation	14	7.1%
Performance Improvement	74	37.4%
Product Development	48	24.2%
Nonspecific	62	31.3%
Total	198	100.0%
Experience
1–2 y	33	16.7%
3–4 y	44	22.2%
5–7 y	70	35.4%
8–10 y	11	5.6%
11–15 y	2	1.0%
Not listed	38	19.2%
Total	198	100.0%
States
California	36	18.2%
New York	22	11.1%
Illinois	16	8.1%
Massachusetts	11	5.6%
Missouri	9	4.5%
Other States	104	52.5%
Total	198	100.0%

There were 4 job-level categories (as noted in Table 1). About two-thirds of the job postings were in the data scientist category and about one-third were classified as senior data scientists including principal and lead roles. Data scientist–level jobs were predominantly found in vendor organizations, health systems, and insurance companies. The minimum required degree was most often a bachelor’s degree. Senior data scientist–level positions were found primarily at vendors and insurance companies. More years of experience was typically required for senior level positions, with 5–7 years of experience required in half of these positions. Higher levels of education were also required, specifically at the master’s degree level.

Positions most often required a bachelor degree and some required a master’s degree. Specific degree areas included quantitative fields such as computer science, engineering, and statistics. The main job focus areas of the posted positions were performance improvement, product development, and innovation. However, about one-third of the job postings were nonspecific. Job postings were generally distributed proportionally across the U.S. The top states based on the number of job postings were California, New York, Illinois, Massachusetts, and Missouri (see Table 1).

Positions varied by hiring organization type. Positions in health systems tended to focus on performance improvement, while vendor positions focused more on product development. Data scientist positions at health systems were found in departments such as enterprise analytics, clinical strategy, informatics, or population health and at insurance companies in departments named clinical analytics or corporate analytics. More advanced health systems or those associated with academic medical centers had data science or artificial intelligence departments. Data scientist positions at insurance companies often did not have a specific job focus and generally supported broader data science needs.

A distribution of top skills based on the job postings is shown in Table 2. Top required skills were statistics, R, machine learning, storytelling (eg, communicating what the data analysis means in a compelling way), and Python (see Table 2). For data scientist–level roles, statistics, storytelling, and R were top skills. For senior-level roles, machine learning, R, and Python were top skills. Microsoft SQL Server, Oracle, and MySQL were top requested relational database skills (see Figure 2 ). R, Python, and SQL were the top programming languages to know. Tableau was the top visualization tool; Spark and Hive were the top big data management platforms.

Table 2.

U.S. healthcare data scientist top 20 skills overall and by job level

Data Scientist Skills	Overall (All Job Postings) n = 3218	Overall Percentage Distribution	Data Scientist Level n = 2006	Senior Data Scientist Level n = 1094
Statistics (eg, general linear model, analysis of variance)	138	4%	94	40
R	136	4%	87	44
Applying machine learning techniques	133	4%	85	44
Storytelling; delivering actionable results	132	4%	91	38
Python	125	4%	79	41
Communicating findings	117	4%	79	36
Developing products	117	4%	75	39
Data-driven problem solving	112	3%	71	39
Data manipulation	108	3%	70	34
Developing algorithms	106	3%	66	34
Setting up/maintaining data platforms	97	3%	65	28
SQL	95	3%	63	27
Implementing models into production	89	3%	54	33
SAS	84	3%	54	27
Work in multidisciplinary teams	84	3%	55	28
Creating visualizations	73	2%	45	25
Identifying business problems to address	64	2%	38	23
Big and Distributed Data	61	2%	34	24
Hadoop	61	2%	25	34
Unstructured Data (eg, noSQL, text mining)	56	2%	34	20
Other	1230	38%	742	436
TOTAL	3218	100%	2006	1094

Figure 2.

Data Scientist Technical Skills.

Job focus area analysis

Four main job focus profiles were developed: performance improvers, product developers, modelers, and innovators. Each is described in this section (see Table 3 for a list of top skills and Figure 3 for a data map by job focus area).

Table 3.

U.S. healthcare data scientist top 20 skills by job focus area

Data Scientist Skills	Overall (All Job Postings) n = 3218	Performance Improvers n = 1181	Percentage Distribution	Product Developers n = 775	Percentage Distribution	Modelers n = 1084	Percentage Distribution	Innovators n = 178	Percentage Distribution
Statistics (eg, general linear model, analysis of variance)	138	55	4.7%	30	3.9%	46	4.2%	7	3.9%
R	136	50	4.2%	32	4.1%	46	4.2%		0.0%
Applying machine learning techniques	133	44	3.7%	41	5.3%	40	3.7%	8	4.5%
Storytelling; delivering actionable results	132	55	4.7%	27	3.5%	40	3.7%	10	5.6%
Python	125	42	3.6%	41	5.3%	39	3.6%	3	1.7%
Communicating findings	117	51	4.3%	18	2.3%	38	3.5%	10	5.6%
Developing products	117	42	3.6%	30	3.9%	36	3.3%	9	5.1%
Data-driven problem solving	112	40	3.4%	25	3.2%	41	3.8%	6	3.4%
Data manipulation	108	47	4.0%	19	2.5%	38	3.5%	4	2.2%
Developing algorithms	106	39	3.3%	27	3.5%	36	3.3%	4	2.2%
Setting up/maintaining data platforms	97	37	3.1%	20	2.6%	34	3.1%	6	3.4%
SQL	95	40	3.4%	23	3.0%	30	2.8%	2	1.1%
Implementing models into production	89	32	2.7%	19	2.5%	35	3.2%	3	1.7%
SAS	84	36	3.0%	6	0.8%	37	3.4%	5	2.8%
Work in multidisciplinary teams	84	32	2.7%	22	2.8%	23	2.1%	7	3.9%
Creating visualizations	73	31	2.6%	15	1.9%	25	2.3%	2	1.1%
Identifying business problems to address	64	26	2.2%	11	1.4%	23	2.1%	4	2.2%
Big and Distributed Data	61	17	1.4%	13	1.7%	26	2.4%	5	2.8%
Hadoop	61	18	1.5%	16	2.1%	25	2.3%	2	1.1%
Unstructured Data (eg, noSQL, text mining)	56	21	1.8%	12	1.5%	18	1.7%	5	2.8%
Other	2240	805	68.2%	572	73.8%	734	67.7%	129	72.5%
TOTAL	3218	1181	100.0%	775	100.0%	1084	100.0%	178	100.0%

Figure 3.

Data Scientist Job Focus Data Map.

Performance Improvers

Performance improvers were sought to work on areas such as quality measures, financial performance, and patient outcomes. In some cases, these positions were specific to areas such as population health, decision support, or biomedical informatics. About two-thirds of these positions were at the data scientist job level vs senior level. Five to 7 years of experience was most commonly required. A bachelor’s degree was required about half the time, followed by a master’s degree. These positions were most often found in health systems or insurance companies. Top skills required included statistics, storytelling, and communicating findings (see Table 3). Examples of key responsibility areas included the following:

“Compile and analyze data to improve health outcomes and increase access to health services for populations at risk.”

“Responsible for defining, building, and improving statistical models to improve business processes and outcomes in one or more healthcare domains such as Clinical, Enrollment, Claims, and Finance.”

“Develops advanced statistical models to predict, quantify or forecast various operational and performance metrics in multiple healthcare domains.”

Product Developers

Product developers focused on a wide range of product development areas including population health, performance improvement, digital health, decision support, speech/language solutions, behavioral health, and claims analytics. The majority of these positions were at the data scientist level and most frequently required 5–7 years of experience. A higher educational level was required, specifically, at the master’s/PhD level. These positions were overwhelmingly found at vendors. Top skills required included machine learning, Python, and R (see Table 3). Examples of key responsibility areas included the following:

“Apply data mining and machine learning techniques to develop better personalization and recommendation for patients’ and doctors’ needs.”

“Develop analyses and user-centered software products that inform the day-to-day decision-making of physicians and hospitals.”

“Develop models to infer or predict disease conditions with a high level of accuracy.”

Modelers

Modeler job postings were nonspecific and required core data science skills. Organizations hiring for these positions were looking for data science bench strength, often machine learning pros. These positions were primarily at the data scientist level with some at a more senior level. Modeler positions typically required 5–7 years of experience. This group of positions most commonly required a bachelor degree. Organizations hiring modelers included insurance companies, vendors, and consulting organizations. Top skills required included R, statistics, data-driven problem solving, and machine learning (see Table 3). Examples of modeler key responsibility areas included the following:

“Serve as the ‘methodology-expert’ on the team and provide guidelinesto team for applying appropriate algorithms or solutions to tackle different problems. Coach junior team members on various data science techniques.”

“Conduct statistical modeling and experimental design on a variety of healthcare datasets, including claims and pharmacy data, biometric data, and healthcare outcomes. Train and validate predictive and categorical algorithms from large datasets.”

Innovators

Innovators were an eclectic group addressing topical areas such as health standards, personalized or precision medicine, genomics, and biology but from a healthcare delivery or informatics perspective. The majority of innovators were at the data scientist level and required 5–7 years of experience. Educational degree requirements varied, with bachelor’s and master’s/PhD being most common. Top hiring organizations included biotech, vendors, and recruiters. Top skills required included communicating findings, storytelling, and product development (see Table 3). Examples of key responsibility areas included the following:

“Working within the innovation team, address real-world challenges such as the rising cost of healthcare, mobility in urban society and payment card fraud in financial welfare programs, using the power of Data Analytics.”

“Develop evidence generation strategies, identify evidence gaps and data sources, design and execute studies, and implement analyses to address molecule and disease area questions; develop next-generation AI for precision healthcare.”

“Design and implement efficient systems for processing large-scale biomedical datasets and conducting rapid experimentation of machine learning and NLP systems.”

DISCUSSION

This study, which was based on job postings, was the first to identify and analyze the required qualifications, skills, and job focus for healthcare data scientist positions. Of note, these positions are often tied to strategic initiatives such as performance improvement, innovation, and product development. The focus area of job postings varied by type of hiring organization, with health system roles generally more focused on performance improvement, and not surprisingly, vendor roles more focused on product development.

There is a spectrum of healthcare data scientist positions. Data scientist roles at health systems and insurance companies are evolving as organizations seek to grow the depth of their operational analytics work. This study found that product developers and modelers have more data science domain expertise and can apply those skills to a wider range of areas. At the more advanced end of the spectrum are the innovators, who have more ability and training in research and science, in particular, precision medicine.

Healthcare organizations are investing in data scientist positions, with vendors and health systems seeking the most applicants. Vendors and insurance companies advertised the most for positions at the senior level, signifying an advancing level of work. With respect to education, a bachelor’s degree was frequently requested though broader education and/or experience was often preferred. As data science expertise can be gained through a variety of channels, hiring requirements must be somewhat flexible.

Healthcare data scientist job skills generally parallel those outlined in previous research. For example, expertise with Spark and Hive were top requested skills for healthcare data scientists which was also the case in previous cross-industry data scientist surveys. With respect to relational databases, Microsoft SQL server was more frequently a required skill vs MySQL and with respect to programming languages, R was more frequently required than SQL was. Finally, machine learning skills were generally requested for healthcare data scientist positions but the applications were not usually specified.

The mix of skills and work varies by job level. For all healthcare data scientist levels, a strong understanding of statistics was a given; whereas for more senior or broader roles proficiency in machine learning was an absolute requirement. This indicates that those individuals interested in a career in data science will need to gain expertise in programming languages such as R and Python. Data scientist–level positions were more closely aligned to analytics across a range of organizational areas. As noted, senior-level positions tended to require stronger programming and modeling skills, as evidenced by the higher ranking of Python and machine learning. These senior-level data scientists often mentor other team members.

While data scientist level roles require technical and statistical expertise, the importance of data communication should not be underestimated. For example, storytelling and communicating findings were top required skills for many positions. Working effectively within a multidisciplinary team was also often listed as a common requirement, signifying the broad base of the work. Also, often listed were proven problem-solving skills, as this ensures the candidate can deal with the challenges involved in addressing business problems in creative ways. Exposure to real-world problems helps develop data acumen.

As shown, data scientist positions require diverse qualifications and skills. Thus, organizations will need to consider candidates with a variety of backgrounds and experience. Critical to filling these positions will be prioritization of the top skills and domain expertise required as well understanding what skills may be developed internally or through additional training. Building a well-defined understanding of talent needs will enable organizations to invest strategically in data science talent pipeline development.¹¹ With this in mind, mentors can help grow talent within an organization. Importantly, organizational culture must fully appreciate this area to ensure retention of key data scientist roles.³⁷

To help meet market demand, analytic skills need to be developed in the broader workforce and career paths need to be provided for analytic professionals to grow their skills over time. Further, educational programs should identify opportunities for students to develop and then apply data science skills. Additionally, as the increase in data impacts nearly every career path, data literacy needs to become institutionalized; moreover, students need to be exposed to data science as early as possible.¹¹

Four limitations were identified in this study. First, the job posting data were collected at only 1 point in time and, as job requirements tend to change over time, this limits the transferability of the findings. Second, as a convenience sample was used, the results are not generalizable. Third, because the content analysis coding was based on the interpretation of 1 researcher, there is a potential for bias. With respect to this third limitation, coding was reviewed and validated multiple times. Last, content analysis is dependent on job posting content which varied and, in some cases, postings were missing key items such as required skills or education. These missing data may impact dependability.

CONCLUSION

This study highlights the critical role of data scientists and how healthcare organizations are seeking data scientists to address specific priority areas. There is no “1-size-fits-all” data scientist, rather the positions may vary broadly. Growing demand for healthcare data scientists requires enhanced professional development, training, and education. The necessary skills for data scientists will likely evolve as market needs change. Further study is needed to monitor trends over time.

AUTHOR CONTRIBUTORS

MAM was the sole contributor.

Conflict of interest statement. None declared.