Recruiting for women's health research: the complete guide

Recruiting for women's health research is harder than recruiting for almost anything else, and the reason is structural rather than accidental. The people you need are defined by conditions that are under-diagnosed, by treatments that are prescribed inconsistently, and by biology that changes week to week. General research panels were not built to hold any of that.

This guide covers what actually determines whether your study fills: your sampling frame, your screening criteria, your incentive, your ethics paperwork, and your timeline. It links to more detailed pieces on each.

Key points

• Define the sample before choosing a recruitment channel.
• Use screening questions that participants can answer reliably.
• Budget for fair incentives, ethics review and realistic fill times.

Why women's health recruitment fails

Most failed recruitment looks the same from the outside. The study goes live, the responses come in fast, and then most of them are screened out. Fill rate collapses and the timeline slips.

Three causes account for most of it.

The sampling frame is wrong. You are recruiting from a general population and hoping enough of them happen to have endometriosis, or be perimenopausal, or be six months postpartum. If the condition affects one in ten women, and your other criteria cut that further, you are paying to screen out ninety percent of the people who click.

The screening criteria were written for a clinician, not a participant. "Confirmed diagnosis of PCOS by Rotterdam criteria" is precise and unanswerable by the person filling in your form. Many women with the condition do not know which criteria were used. Some were diagnosed years ago by a GP who never named a framework.

The incentive does not match the ask. A twenty minute survey about a painful and private subject is not the same as a twenty minute survey about supermarket brands, and it should not pay the same.

Choosing a sampling frame

There are four realistic routes, and each has a different cost and time profile.

1. Clinical sites. Highest data quality, slowest, most expensive, and dependent on clinician goodwill. Right for interventional work.
2. Patient advocacy groups and charities. Excellent reach into specific conditions and high trust, but the sample skews towards the most engaged and most severely affected.
3. General research panels. Fast and cheap per response, but health screening happens at the door, so you pay for the screen-outs and you cannot verify much.
4. Condition-screened panels. The pool is screened on health before your study exists, so matching is a lookup rather than a hunt.

FeFe is the fourth kind. That is a disclosure, not an argument. There are studies where a clinical site is the only defensible choice, and we say so in FeFe compared with general research panels.

Writing screening criteria people can answer

Translate every clinical criterion into something a participant can answer about her own life without a medical record in front of her.

• Instead of asking for a diagnostic framework, ask who diagnosed her and roughly when.
• Instead of asking for a drug class, list the actual product names she would recognise from a box or a prescription.
• For cycle-dependent studies, ask for the date of the start of her last period, not for a cycle phase label.
• Give an honest "I am not sure" option. It is more useful than a confident wrong answer, and it lets you route the uncertain cases to a follow-up question rather than losing them.

Then write the exclusions separately, and keep them short. Every extra exclusion multiplies through the funnel.

Setting the incentive

Work out the real completion time by piloting the task with three or four people, not by estimating it. Then consider paying at or above the independently calculated real Living Wage for that time, rounded up. [1]

Two adjustments are worth making in women's health specifically. Pay more for tasks that ask about trauma, loss, fertility or pain, because the emotional load is part of the work. And pay for screening time if your screener is longer than a couple of minutes.

Ethics and data protection

If your study needs review, recruitment is part of what gets reviewed. Ethics review considers recruitment, consent, payment, confidentiality and withdrawal arrangements as part of protecting participants. [2]

Health data is special category data under UK GDPR, so processing requires an Article 9 condition as well as an Article 6 lawful basis. Scientific research and explicit consent are possible Article 9 conditions, but the correct route depends on the organisation and study. [3]

Realistic timelines

Be honest with your stakeholders about what recruitment speed depends on. A screened pool removes the search time, not the study design time. Broad criteria and a fair incentive fill fast. Narrow criteria fill slowly wherever you recruit, because the people simply are not numerous.

Where to go next

• How to write screening criteria for menopause studies
• How much to pay research participants in the UK
• UK GDPR and health data in research recruitment

If you want to see how a condition-screened pool changes the timeline for your specific study, you can set up a study for free and see the matching pool before you spend anything.

References

1. Living Wage Foundation. What is the real Living Wage?
2. Health Research Authority. Applying to a Research Ethics Committee.
3. Information Commissioner’s Office. What are the conditions for processing special category data?