Running a survey of the profession
A survey of medical communicators needs clear objectives, an appropriate questionnaire and recruitment plan, protected responses, and analysis that respects the sample’s limits.
Objectives and questionnaire development
Objectives and the target audience should be settled before questions are written, because collecting questions first can produce an unwieldy instrument. In a 2021 veterinary-writing survey, clarifying the objective as defining veterinary medical writers led to seven questions: work areas, professional paths, employment status, location, career goals, ranked meeting topics and free comments. The initial questionnaire profiled people working in or interested in veterinary writing, while a narrower follow-up targeted established writers.
AMWA’s 2013 needs assessment focused on educational programmes, services and resources rather than member satisfaction alone, and was developed by senior staff and its education and member-resources departments.
A 2016 survey of pharmaceutical and device writers combined multiple-choice questions, Likert scales and free text, permitting multiple answers where appropriate; senior writers and managers tested it before revision. AMWA and CMR’s 2009 industry questionnaire had 59 questions in eight categories: organisational structures, products and services, outsourcing, resources, education and qualifications, productivity, quality, and cycle time/review cycles. Seven key opinion leaders drawn from AMWA membership and representative of the target population reviewed the categories and draft questions; team testing preceded programmers’ final corrections. Multiple-choice questions included free text for “other”, and the interface used an existing survey shell while questions were finalised.
Italy’s 2024 questionnaire built on recent EMWA surveys and used a custom platform to present questions according to employment type. Its four sections covered demographic, educational and professional characteristics; employee/hybrid responsibilities and salary; freelance taxation, clients, outsourcing, income, payment and services; and agency taxation, staff, clients, income, payment and services. Most answers were fixed categories or Likert scales, with free text for matters such as tax codes, income and numbers of employees or freelances used.
EMWA’s 2018 and 2023 freelance surveys were designed by its Freelance Business Group committee to take under ten minutes, increasing from 18 questions to 35. Both covered demographics, work, pricing/payment, business and networking, with wellbeing added in 2023. Fixed categories, Likert scales, free text and mixed responses were used, generally allowing an “other” answer.
AMWA’s 2015 survey task force streamlined the previous questions and produced a preparation guide, asking respondents to have a W-2 or profit-and-loss statement available; with the guide, completion was expected to take about ten minutes.
Eligibility and definitions
AMWA’s 2019 compensation survey admitted anyone who had worked for pay as a medical communicator in 2018, full- or part-time. Employees had taxes deducted by employers; freelances paid their own taxes, including contractors, consultants and small-business owners paying themselves a salary. Full-time freelance work meant at least 32 weekly hours, including billable and nonbillable time, to align with earlier rounds. Freelances charging by project converted to hourly rates using the time normally needed for a project.
The 2007 AMWA salary survey specified these definitions:
- Employees worked for a company, institution or individual, with employer-paid Social Security and deducted state and federal taxes.
- Freelances worked for clients, paid their own Social Security and had no client deduction of state or federal taxes.
- Full-time employment meant at least 32 hours weekly; part-time meant fewer than 32.
- Employee total income was before tax deductions.
- Freelance gross income was all money collected from clients.
- Freelance net income deducted expenses such as insurance, subcontracting and equipment, before taxes and retirement contributions.
The 2015 survey screened for actively working full- or part-time medical communicators, requested 2014 tax-year income and directed responses by employee or freelance status.
Administration and funding
For the 2009 industry survey, AMWA partnered with CMR, a pharmaceutical and biotechnology research firm, using existing survey and reporting technology to study medical-communication managers. With no association funding, participating companies paid CMR for customised comparisons against blinded participant data, while AMWA team members and expert reviewers volunteered without payment. Recruitment through AMWA and CMR’s existing clients favoured large companies and created potential bias.
For 2019, AMWA commissioned Association Research Inc to administer its compensation survey, manage analysis and prepare the report, with a volunteer task force shaping questions and analyses. The partnership extended recruitment beyond members and enabled faster reporting. In 2019, AMWA described its use of historical, aggregated compensation data and an independent contractor as measures to mitigate antitrust concerns and protect individual privacy.
Employment and compensation questions
EMWA’s 2021 compensation instrument defined salaried employment as company, institutional or individual employment, including full-time, part-time and fixed-term work, with taxes paid on the respondent’s behalf and payment by salary or hourly wage. Its questions included the following:
| Question | Response structure |
|---|---|
| Sex | Male, female, non-binary, prefer not to say. |
| Country of tax residence | Country list or prefer not to say. |
| English as native language | Yes, no, prefer not to say. |
| Highest degree | Associate or below, bachelor’s, master’s, advanced medical/doctoral/professional degree or equivalent. |
| Field of highest degree | Life sciences, healthcare, applied sciences, humanities, languages/translation, other. |
| EMWA certificate | Yes or no. |
| Other formal medical-writing training/certification | Yes with details or no. |
| Medical-communication experience | Up to 2, 3–5, 6–10, 11–15, over 15 years. |
| Prior work in science | Yes or no. |
| Employment type | Employed, freelance, hybrid, small-business owner with fewer than ten salaried or subcontracted team members; multiple answers allowed. |
| Employer or main-client type | Pharmaceutical, veterinary, medical device, biotech, medical-communications/advertising/consultancy, CRO, academia/government/not-for-profit, other. |
| Job title | Associate, junior, senior or principal medical writer; medical-writing manager; communication lead/head; publishing specialist; director/company owner; freelance; other; multiple answers allowed. |
| Project management | Yes or no, defined as project oversight without line management. |
| Mentoring | Yes or no. |
| Line management | Yes or no. |
| Weekly hours | 1–10, 11–20, 21–30, 31–40, 41–50, over 50; freelances average across projects. |
| Work location | Fully office-based, fully remote, office with home days, remote with office days. |
| Primary role | Medical-communications writing, regulatory writing, other writing, editing, translation, management, other; working team leaders select their team’s category. |
| Income | Employee or hybrid pretax annual income to the nearest €1,000; freelance pretax hourly rate to the nearest €10, converted at the valid exchange rate. |
| Workload percentages | Drug-development documents, device-development documents, journal/scientific-press articles, marketing/congress materials, patient/professional education, grants, translation, social media/electronic publishing, other. |
| Work and salary satisfaction | Separate five-point scales from very satisfied to very unsatisfied. |
| Reasons for satisfaction level | Excessive workload, insufficient challenge, unsupportive environment, discrimination, undervalued role, low salary, difficulty finding regular clients, none/other; multiple selections allowed. |
| Pandemic effects | Workload fluctuations, tighter deadlines, workspace disruption, unemployment/fewer clients, health/family-related inability to work, no disruptions, increased activity, reduced face-to-face contact, other; multiple selections allowed. |
Freelance business questions
The 2023 freelance questionnaire contained these questions and response choices:
| Question | Response structure |
|---|---|
| 1. Gender | Female, male, non-binary, prefer not to say. |
| 2. Age | Under 20, 20–29, 30–39, 40–49, 50–59, 60–69, at least 70, prefer not to say. |
| 3. Base | Europe or outside Europe, with country specified. |
| 4. Setting up freelance | Whether straightforward; if not, issues faced in free text. |
| 5. Occupational status | Entirely freelance, hybrid freelance/contract/student, other. |
| 6. Freelance experience | Under 1, 1–5, 6–10, 11–15, over 15 years. |
| 7. Weekly freelance hours | 1–10, 11–20, 21–30, 31–40, 41–50, over 50. |
| 8. Services and time share | Writing, proofreading/editing, translation, quality control, electronic publishing, consultancy such as training, other; select all and give percentages. |
| 9. Documents | Journal manuscripts; conference abstracts/posters/slides; marketing; market-access/HEOR; proceedings; educational/training materials; textbooks; regulatory protocols/reports/brochures; summaries/CTD/SmPCs/inserts; device/biologic regulatory documents; other regulatory documents; SOPs/manuals; websites/online articles; other; select all. |
| 10. Success skills | Project management/delivery, marketing, writing, website design, accountancy, networking, statistics, work–life balance, other; select all. |
| 11. Quote preparation | Estimated time multiplied by a time rate; time rate with all time invoiced afterwards; document/project price list; client/location price list; per-word/page translation fees; other. |
| 12. Typical job hours | Writing, editing or translating a standard scientific article; CSR/other regulatory writing; technical evaluations; medical blogs; patient materials; website content; other. |
| 13. Hourly rates in euros | Regulatory writing, non-regulatory writing, editing, proofreading, translation, quality control, consultancy, training. |
| 14. Milestones | Percentage upfront, interim payments, non-refundable future-booking deposit, combination, other, no milestones. |
| 15. Minimum monthly earnings | Free-text amount in euros. |
| 16. Tax/insurance fixed costs | Printed bands: over 20%, 20–30%, 30–40%, 40–50%, over 50%. |
| 17. Payment recovery | Emails/reminders, client call, third-party collection, other, never had issues. |
| 18. Major project source | One answer: longstanding clients, colleague referrals, client referrals, professional/social networks, own website, freelance directory, other. |
| 19. Clients | Pharmaceutical; biotech/veterinary/device; communications agencies; writing companies; CROs; academia; publishers/health journalism; nonprofits; other; multiple selections. |
| 20. Advertising | Website, social media, directories, associations/societies, print, prospective-client contact, none, other; select all. |
| 21. Main influence on new business | One answer: website, social media/LinkedIn, professional platforms, client word of mouth, EMWA directory, other. |
| 22. Development | EMWA workshops, webinars, other-association workshops, online workshops, personal research, other; select all. |
| 23. Networking | EMWA or other-association conferences, company courses, local meetings, professional/social networks, none, other; select all. |
| 24. Associations | EMWA, AMWA, Australasian association, ISMPP, other, none. |
| 25. EMWA membership duration | Under 1, 1–3, 4–5, over 5 years; members only. |
| 26. EMWA conference attendance over three years | 1–2, 3–4, 5–6, none; members only. |
| 27. Most helpful EMWA activities | Association, workshops, freelance forum, networking, webinars, symposia, expert seminars, SIGs, journal, other; members select all. |
| 28. Directory | Whether listed; those listed rate its usefulness for advertising and prospective-client contact using the printed agreement options from strongly agree to somewhat disagree. |
| 29. Future membership | Non-members state whether they plan to join. |
| 30. Additional activities | No or yes with suggestions. |
| 31. Workplace | Own office, shared space, client site, home, office/home hybrid, other. |
| 32. Personal work stress | Never, sometimes, most times, always. |
| 33. Stress triggers | Workload/deadlines, finances, COVID issues, retaining/losing clients, finding clients, work–life balance, other; select all. |
| 34. Stress reduction | Whether actively attempted, and methods including physical activity, mindfulness/meditation, social activities, home activities, other. |
| 35. Outreach | Free-text views and suggestions about EMWA’s support for freelance writers. |
The closing message promised publication of full results on EMWA’s website and in its journal.
The printed 2018 questionnaire used the following shorter sequence:
| Question | Response structure |
|---|---|
| 1. Work status | Part-time or full-time freelance. |
| 2. EMWA membership | Yes/no; non-members asked whether they plan to join. |
| 3. Base | Europe or outside Europe, with country. |
| 4. Weekly hours | 1–10, 11–20, 21–30, 31–40, 41–50, over 50. |
| 5. Services/time share | Writing, proofreading/editing, translation, quality control, electronic publishing, consultancy/training, other; percentages and “select all that apply”. |
| 6. Documents | Journal manuscripts; abstracts/posters/slides; marketing; proceedings; education/training; textbooks; regulatory protocols/reports/brochures; summaries/CTD/SmPCs/inserts; devices/biologics; other regulatory documents; SOPs/manuals; websites/online articles; other; “select all that apply”. |
| 7. Charging basis | Hourly, daily, weekly, monthly, project, other. |
| 8. Euro hourly rate | Regulatory and non-regulatory writing, editing, proofreading, translation, quality control, consultancy, training. |
| 9. Milestone frequency | Never, sometimes, most times, always for long-term projects. |
| 10. Payment problems | Never, sometimes, most times, always. |
| 11. Major project source | Longstanding customers, colleague/customer referrals, professional/social networks, own website, freelance listings, other. |
| 12. Main clients | Pharmaceutical; biotech/veterinary/device; communications agencies; writing companies; CROs; academia; publishers/health journalism; nonprofits; other. |
| 13. Advertising | Website, social media, directories, print, prospective-client contact, other; select all. |
| 14. Directory | Members state whether listed; listed members rate usefulness for prospective-client contact; unlisted members rate willingness to list if given more information, on strongly-agree to strongly-disagree scales. |
| 15. Development | EMWA workshops, other-association workshops, online workshops, other; select all. |
| 16. Networking | EMWA and other-association conferences, company courses, local meetings, professional/social networks, other; select all. |
| 17. Conference attendance | Members report 1–2, 3–4, 5–6 or no EMWA conferences in three years. |
| 18. Helpful EMWA activities | Association, workshops, freelance forum, networking, webinars, symposia, expert seminars, SIGs, journal, other; members select all. |
| Suggestions | Free-text views and ideas about EMWA’s freelance outreach, intended to guide new benefits and initiatives. |
The 2015 AMWA survey retained sex, age, education, experience, work status, primary employer, work type and career level for trend analysis, rephrasing and reordering some questions for accuracy.
Its new questions covered:
| Question | Response structure |
|---|---|
| All respondents | Percentage working remotely: none, 20%, 40%, 60%, 80%, all. |
| Employees | Employer size: under 50, 50–100, 101–499, 500–1,000, 1,001–4,999, 5,000–10,000, over 10,000; years in current post; twenty benefit choices plus none/other. |
| Editors | Main level: macroediting, microediting, both, copyediting. |
| Freelances | Business structure: LLC, partnership, S corporation, sole proprietor, other; recurring expenses with ten options and other. |
| Freelance preparation | Years of medical communication before full-time freelancing: none, under 1, 1–2, over 2 but under 5, 5 to under 10, at least 10. |
| Freelance service mix | Time percentages for the three leading service areas. |
| Subcontracting | Often, sometimes, never. |
| Marketing | Work-time bands below 10%, 11–19%, 20–30%, over 30%, and percentage of annual revenue spent. |
| Retirement | No contribution, maximum permitted contribution or a lower contribution in 2014. |
Recruitment and response
The 2016 pharmaceutical/device survey sent a covering letter and link to the authors’ contacts in the USA and Europe, inviting onward circulation to industry, agency and freelance colleagues and aiming for 50 completions. It ran from 8 to 14 March 2016 and received 112 responses from an estimated 200 contacts, reported as approximately 56%.
The 2021 veterinary survey used its special-interest group’s mailing list, EMWA’s newsletter and social media, inviting onward sharing among relevant colleagues. Repeated reminders were needed before responses reached double figures, and the final total was 34.
Recruitment for the 2009 industry survey included August and September member-newsletter announcements, a website notice, an October conference booth, the contractor’s contacts in the USA, Europe and Japan, a prospectus and telephone/personal follow-up from August to October. Data collection ran in November and December 2009.
AMWA’s June 2013 needs assessment invited approximately 5,000 current members and 2,000 people whose membership had lapsed within three years, supplemented by LinkedIn, Facebook and Twitter invitations. Of 1,043 starters, 944 completed it, or 91%; 17% of members participated.
EMWA’s freelance surveys invited all freelance medical writers, including non-members, through subscriber emails, its website, social media and authors’ professional networks. The 2018 round also used American and Australasian association networks; it ran from 19 December 2018 to 1 February 2019 on Google Forms, while the 2023 round ran from 13 to 31 January on SurveyMonkey, both in English.
Italy’s survey invited writers living in Italy and Italian writers abroad regardless of level, using EMWA channels, direct emails, social media and word of mouth throughout February 2024. It drew 68 responses compared with no more than eight Italian responses in earlier EMWA surveys, attributed to local-group engagement, personal involvement in its LinkedIn community and executive-committee support.
AMWA’s 2019 survey used individual links and offered early results plus a draw for three $100 gift cards. A login allowed respondents to pause and return; invitations advised safe-listing the sender and promised summary results to participants and expanded results to members. Non-member communicators could request invitations. There were 1,418 respondents, a reported 20% response rate, including 480 freelances and 179 full-time freelances.
In 2015, AMWA recruited non-members for the first time, including lapsed members, sister associations and LinkedIn groups, with reminders, journal and postcard announcements. The launch notice gave an April 24 deadline, while the methods report records availability from April 7 to May 20. Recommendations after the 2011 survey included longer response periods, postcard reminders, emphasis on participation, social-media promotion and explicit opt-outs to distinguish them from nonresponse.
Confidentiality and access
The 2016 pharmaceutical/device survey involved no patients, sought no ethics-committee approval, offered no inducement or compulsion and allowed questions to be skipped. AMWA’s 2013 needs assessment was anonymous. EMWA’s 2023 freelance and Italy’s 2024 surveys collected anonymous data under GDPR and treated voluntary participation as consent.
In the 2009 industry survey, secure online responses were stored on CMR servers; analysts could identify respondents, but other participants saw blinded data and AMWA received aggregated data without individual or company identities. The database was locked at closure and analysed in SQL Server; an online tool let participants compare their answers with the dataset. Association Research promised confidentiality for individual 2019 compensation responses.
The AMWA-Canada survey assigned unlinked identifiers to emailed questionnaires, stored them in a folder for committee analysis and deleted covering emails, reporting only summaries to avoid identifying individuals. EMWA’s 2021 salary questionnaire allowed confidentiality-sensitive questions to be skipped and promised no publication of individual responses.
Cleaning, analysis and reporting
AMWA’s 2013 needs assessment calculated percentages from each item’s respondents, with at least 900 responses for most questions and separate counts for smaller items. In 2007, required questions still had different response totals because exiting the instrument left subsequent answers blank.
The 2016 pharmaceutical/device survey reported counts, percentages, means and quotations, without statistical testing because it was exploratory and possibly unrepresentative. In the 2019 freelance analysis, a sample with 47% aged at least 55 and 73% having over ten years’ experience led to reporting medians and interquartile ranges, which are less sensitive to outliers.
Italy’s 2024 survey exported responses to Excel, standardised free-text values, treated null or inappropriate entries as null and excluded them, then used Power BI for medians, minima/maxima, averages, percentages and graphics without inferential analysis.
EMWA’s 2023 freelance analysis exported to Excel, coded free-text responses by theme and merged poorly represented categories, below 5%, into age groups below 39, 40–49, 50–59 and at least 60, and weekly-hours groups below 20, 20–40 and above 40. Different age bands required estimates from divided 2018 response groups for comparison. Writer type was assigned from documents delivered, and each respondent’s average hourly rate was the mean across their service-specific rates. Stata 17 produced means and frequencies, using all respondents unless a subgroup denominator was specified and noting omitted answers. Tests included Student’s t, chi-square, Fisher’s test for contingency values below five and one-way ANOVA, with P<0.05 as the significance threshold.
AMWA’s 2007 survey consulted a statistician, used SAS for descriptive statistics, regression with interactions for full-time employee salary predictors and Pearson correlations for hours versus income and work versus pay satisfaction. Its regression considered education, sex, age, experience and three employer groups: pharmaceutical/biotechnology companies; device, communications or advertising companies; and other employers. The 2007 questionnaire requested salary to the nearest $1,000, sex, age, education and degree field, location, work status, employer, career level, experience and work/pay satisfaction. It was open to members from April 18 to May 12, asking about 2006 income, after a journal notice and member email, a launch email and a reminder ten days later. In 2011, an independent adjudicator with experience of three earlier surveys reviewed questionable responses.
For 2015, raw responses were preserved in Excel, imported into SAS and cleaned programmatically, with approximately 100 uncorrectable entries adjudicated by an author. Descriptive analyses covered all questions; small response categories were grouped and freelances compared by full/part-time status, writing/editing work and marketed specialism. Employee salary regressions considered sex, age, education, experience, employer and geographic region; initial models were optimised for statistical significance and correlation before exploratory regressions. US regions were grouped by low, medium and high composite consumer price index. Means were generally accompanied by sample size, standard deviation and median, sometimes a range, because distributions were often variable and skewed, especially for small samples.
Limitations and comparisons over time
Several respondents found the 2016 pharmaceutical/device questionnaire too long; replication recommendations were to validate questions, shorten it and recruit a much larger sample. Unanswered questions included the nature, participants and outcomes of discrimination or ethical issues and whether transatlantic salary differences reflected living costs, selection bias or job differences.
The 2019 AMWA freelance findings were descriptive because representativeness of its convenience sample was unclear. Direct statistical comparison with earlier rounds was inappropriate because respondent pools, question wording and presentation could differ, although consistent rates and time-based billing patterns lent credibility to broad trends. Surveys show association rather than causality, and condensed summaries restrict available detail and conclusions.
The small, large-company-oriented 2009 industry sample limited generalisation and did not directly measure the overall share of outsourced work. Broader recruitment was recommended across organisation types, including universities, health centres and agencies, and beyond the USA and Europe.
Italy’s 2024 sample remained small despite improved participation, varied by services, experience and geography and could not support reliable subgroup analysis; its two agency owners appeared only in totals. Voluntary anonymous data could not be validated, and results could not be generalised even to the Italian profession.
EMWA’s 2023 freelance survey over-represented the UK and missed many countries, preventing country-specific conclusions. Network-based distribution could introduce selection bias, low-response subgroups prevented clear regression conclusions and overlap between rounds was unknown. Added questions prevented complete comparisons; writer type and average rate were derived rather than directly asked, experience did not distinguish medical writing from freelancing, and satisfaction and pandemic impact were not asked. Rates and income expectations varied widely, with few responses for some services and no explanation for high monthly expectations, so they were offered only as a guide.
The 2018 freelance questionnaire used frequency scales for milestone payments and payment problems, whereas the 2023 instrument asked payment types and retrieval methods. The report describes three project sources allowed in 2018 versus one in 2023, and three service/document choices in 2018 versus all applicable choices in 2023, although the printed 2018 service/document questions say “select all that apply”. The printed 2018 development, networking and advertising choices lack the explicit personal-research, no-networking and no-advertising options in 2023.
AMWA’s five salary surveys from 1989 to 2007 retained valuable questions for comparisons while adding requested information, including freelance writing and editing questions. The seventh survey in 2015 similarly retained key questions while adding and refining others. EMWA’s 2021 survey admitted members only, so its findings could not be generalised to the wider European or global profession; voluntary participation could create selection bias, living costs were not considered and the interface allowed illogical multiple answers. EMWA established a dedicated salary and compensation team in 2021 to repeat surveys every four to six years and improve their design using feedback.
Uses of the results
AMWA planned to use its 2013 needs assessment to develop, revise or remove educational programmes and services for different roles and career stages and allocate resources within its budget. In the veterinary survey, free comments provided more detail than predefined work categories, and a one-question follow-up asked which species established writers covered. The results informed the group’s 2021–2022 activities. Ranked meeting topics showed no clear central tendency, but an untested pattern linked experienced writers with interest in writing areas and less-experienced writers with careers and training. That pattern informed a varied quarterly speaker/topic programme for 2022 and expanded workshops, including One Health in autumn 2021.
AMWA’s 2019 contractor data enabled a members-only calculator showing means, medians and quartiles by experience, degree and work setting, alongside a report exceeding 200 pages of statistical tables and salary, benefit and income information. The report was intended for sale to employers and companies, with member discounts, supporting competitive compensation packages, salary negotiation and freelance pricing.
AMWA’s 2007 reporting used standalone tables and a figure, an expanded overview table for historical comparison and brief text for qualitative details, smaller-sample findings, methods and discussion. Additional slides supplied pharmaceutical/biotechnology detail and salary algorithms based on education, experience, duties and career level.