Best-practice assessment of patient-reported outcomes in cancer clinical trials

PROMs can:

• tell us whether benefits to survival from treatment outweigh side-effects and costs;
• tell us what supportive care should accompany treatments found to improve survival;
• screen for patient eligibility and provide prognostic information;
• assist regulatory authorities, clinicians and patients to make decisions between different treatments; and
• give patients a 'voice' appropriate to the current era of consumer-lead, value-based healthcare. 

Estimates suggest that 30% of clinical trials have over 20% of missing or invalid quality of life data. 

An example of invalid data is:

Missing and invalid data reduce the statistical power of a trial and introduce 'bias'. 

The following guidance addresses issues raised by research trial staff and is based on guidelines to minimise missing PROM data. 

Mode of administration

Familiarise yourself with the contents of each PROM and mode of administration. This includes whether the PROM will be administered via:

• self-completion or interviews (face-to-face or telephone), and
• pen-and-paper or electronic formats. 

Timing of administration

Determine the timepoint(s) when PROMs need to be administered, for example, at baseline and two-week follow-up. Usually, there will be a 'window' allowed around each timepoint - for example, 1 or 2 days. Timepoints and windows should be defined according to when there are likely to be effects from treatment and disease - for example, at the end of each chemotherapy cycle.

Also determine any requirements for the timing of PROMs within the clinical visit or schedule of other assessments, for example, at the start or end. 

If more than one PROM is being administered, determine if they are to be ordered in a specific way or in a random order each time. Ordering is important because patients' responses on one PROM can sometimes influence their answers to those after. 

reminders for electronic proms

If PROMs will be administered electronically for patient self-report, set regular reminders, for example, via automated text messages.

patient privacy

Where possible, make sure the patient is in a quiet space away from other people because:

• many PROMs include personal or sensitive questions that patients may way to keep private, and
• quiet environments help with concentration for patients facing cognitive challenges from cancer or treatment.

patient motivation

Explain the patients that PROMs are important so they're more likely to complete them with care and attention. 

Talk the patient though the CQUEST patient fact sheet, 'Understanding questionnaires in cancer clinical trials' to help them understand why PROMs are important and the dos/don'ts of completing them. 

should i provide any assistance to the patient in self-completing PROMS?

If the protocol for the trial asks for patients to self-complete PROMs, please offer as little help as possible.

At most, limit your assistance to reading the questions out loud and recording the responses for patients who are unable to do this for themselves.

If patients don’t understand a question or ask for help with it:

• Never reword or try to explain it.
• Instead, ask the patient to answer as best they can. Reassure them that questions can mean different things to different people, and there are no right or wrong answers. 
• If the patient remains unsure how to answer the question, leave the question blank, making a note of the reason why the data are missing.  

check for completeness

The Completion and Missing Data (CoMiDa) Form is recommended to assist with documenting PROM assessment at each timepoint. 

Ensure the front cover of the PROM is completed with any details required, for example: the patient's study ID, the date and time of completion, and which staff member collected the PROM data via interview. 

Wherever possible, check the PROM responses for completeness immediately after completion. Ask the patient to complete any questions they may have missed. 

• Make notes for any reasons a patient has given for not completing a question. 
• If answers appear inconsistent, invite the patient to read over and check their responses. If they choose to stick with the answers they provided, make a note that their responses were confirmed. 

Sometimes PROMs are not received until after the timepoint window specified in the protocol - for example, if patients post them back via snail-mail. In these cases, it will not be possible for patients to change their responses, but you may still investigate reasons for missing data. 

REsponding to PROM data of concern

Some PROMs include questions that require clinical action if a patient answers in a certain way - for example, suicidal ideation or severe pain. These are commonly referred to as PRO alerts.

According to the Good Clinical Practice, the wellbeing of participants needs to be prioritised over the needs of the trial. 

Familiarise yourself with any PRO alert protocols in place for the clinical trial in question. Seek clarification from the study coordinator/investigator team if you are unclear on any aspect. Formally, record any off-protocol interventions administered because of PRO alerts. 

If you are unsure about any aspect of PROM data collection, please refer to the clinical trial protocol. 

For a checklist for clinical research professionals to support the systematic collection of PROM data, please see Wehrlen et al. (2016). ⁸

For guidelines to reduce, handle and report missing data, please see Hussain et al. (2022). ¹⁰

 1.             Au H-J, Ringash J, Brundage M, Palmer M, Richardson H, Meyer RM. Added value of health-related quality of life measurement in cancer clinical trials: the experience of the NCIC CTG. Expert Rev Pharmacoecon Outcomes Res. 2010;10(2):119-128.

2.             Ford JH, Foster SA, Nichols RM, et al. A real-world analysis of patient-reported outcomes in patients with migraine by preventive treatment eligibility status in the US and Europe. Journal of Patient-Reported Outcomes. 2020/07/06 2020;4(1):53. doi:10.1186/s41687-020-00221-w

3.             Mierzynska J, Piccinin C, Pe M, et al. Prognostic value of patient-reported outcomes from international randomised clinical trials on cancer: a systematic review. Lancet Oncol. Dec 2019;20(12):e685-e698. doi:10.1016/s1470-2045(19)30656-4

4.             Schorr A. Patient-reported outcomes and where patient advocacy groups can help. Change Together. 2017;

5.             Fielding S, Ogbuagu A, Sivasubramaniam S, MacLennan G, Ramsay CR. Reporting and dealing with missing quality of life data in RCTs: has the picture changed in the last decade? Qual Life Res. Dec 2016;25(12):2977-2983. doi:10.1007/s11136-016-1411-6

6.             Mercieca-Bebber R, Calvert M, Kyte D, Stockler M, King MT. The administration of patient-reported outcome questionnaires in cancer trials: Interviews with trial coordinators regarding their roles, experiences, challenges and training. Contemp Clin Trials Commun. Mar 2018;9:23-32. doi:10.1016/j.conctc.2017.11.009

7.             Kyte D, Ives J, Draper H, Keeley T, Calvert M. Inconsistencies in Quality of Life Data Collection in Clinical Trials: A Potential Source of Bias? Interviews with Research Nurses and Trialists. PLoS One. 2013;8(10):e76625. doi:10.1371/journal.pone.0076625

8.             Wehrlen L, Krumlauf M, Ness E, Maloof D, Bevans M. Systematic collection of patient reported outcome research data: A checklist for clinical research professionals. Contemp Clin Trials. 2016;48:21-29. doi:10.1016/j.cct.2016.03.005

9.             Kyte D, Draper H, Calvert M. Patient-reported outcome alerts: ethical and logistical considerations in clinical trials. JAMA. Sep 25 2013;310(12):1229-30. doi:10.1001/jama.2013.277222

10.          Hussain JA, White IR, Johnson MJ, et al. Development of guidelines to reduce, handle and report missing data in palliative care trials: A multi-stakeholder modified nominal group technique. Palliat Med. 2022;36(1):59-70. doi:10.1177/02692163211065597