Table 1 Summary of included cases

Case 1: NONSPEX

Single-blind feasibility study assessing the effects of including nonspeech oromotor exercises in SLT rehabilitation for persons with dysarthria following stroke (Home across 6 NHS health boards)

All participants (n = 39): Mins of words and sentence, conversation and (intervention only) nonspeech oromotor exercises per day and number of practices per day (7 weeks)

▪Interview

▪Sample diary

▪Viewed sample of completed diaries

▪Results publication [36]

▪Conference presentation [37]

3 A4 pages (×7); 126 items per week; frequency (tick), duration (number), comments (open); collected weekly

High – 100 % (32/32)

Variable – a few people ‘physically not able to’ (NONSPEX: interview) complete diaries. A wide range of adherence was reported, suggesting variable completion. The random sample viewed in interview had variable completion.

Unclear: ‘the therapist has recorded “some participants may have recorded more practice than they had carried out, some may have recorded less” (NONSPEX: interview)

Case 2: SCORD

Substantive pragmatic open trial assessing the use of night resting thermoplastic splints following hand surgery and therapy for Dupuytren’s contracture (5 NHS trusts, home)

Experimental (n = 77 + per-protocol deviations): Number of nights per week the splint was worn (6 months)

▪Interview

▪Sample diary

▪Viewed sample of completed diaries

▪Results, protocol and survey publications [38–40]

▪Relevant participant information sheet section

1 A4 page (×2); 3 items per week (completed weekly); frequency (number), comments (open); collected trimonthly

High – diaries returned completed in almost all cases (SCORD: telephone discussion). Successfully classified participants for a per-protocol analysis.

High: ‘there might be [missing data]… I did dig out a couple of diaries, just very randomly … none of those had any missing lines’ (SCORD: interview). Random sample viewed in interview were complete.

Unclear: ‘diaries were collected by research associates (not the treating therapists or surgeon) and, therefore, encouraged patients to be honest, independent verification of actual splint wear was not possible’ (SCORD: results publication). ‘What came back in the diaries was good, but it wasn’t too good to be true.’ (SCORD: interview)

Case 3: ENVISAGE-WP2

Pilot open trial (n = 22) comparing exergames, visualisations and exercise booklet only for secondary falls prevention (home only, single area)

All participants (n = 22): Number of individual exercises and exercise sessions undertaken (12 weeks)

▪Interview

▪Sample diary

▪Viewed sample of completed diaries

▪Results and protocol publications [41, 42]

▪Graph comparing computer and diary-recorded adherence

1 A4 page (×12); 0–70 items per week; frequency (tick); collected at end of study

High – 100 % (22/22)

High: ‘there was a lot of detail in the diaries’ (ENVISAGE: interview). Sample of diaries viewed completed well.

High: in the exergame group (n = 7) 7 weeks were correct, 5 weeks varied by <3/27 exercises across all participants.

Case 4: SELF pilot

Pilot pragmatic open trial comparing self-managed loaded PT exercise with usual PT for rotator cuff tendinopathy (single centre, private clinic)

Experimental (n = 12): Completion of sets of exercises per day (1 tick per session for 2 sessions a day) (3 months)

▪Interview

▪Sample diary

▪Viewed sample of completed diaries

▪Pilot results publication [43]

▪In -press results of substantive study

▪Substantive study protocol [44]

▪Qualitative pilot publication [45]

▪Patient and public involvement publication [46]

1 A4 page (×1–5); 7 items per week; frequency (tick); returned at each appointment

High: 92 % (11/12)

Medium-high: 7 complete data (all consecutive diaries), 4 partial data. Returned diaries were well-completed on viewing, with the occasional incorrect completion. ‘reasonable degree of consistency’ (SELF: interview).

Unclear: ‘therapists were very sceptical about this … you can see that sometimes the form and the pen is exactly the same and they’ve obviously been done quickly’. (SELF: interview)

Case 4: SELF substantive

Pragmatic open substantive trial comparing self-managed loaded PT exercise with usual PT for rotator cuff tendinopathy (3 NHS PT departments, home)

Experimental (n = 42): Completion of sets of exercises per day (one tick per session for two sessions a day) (3 months)

1 A4 page (×1–5); 7 items per week; frequency (tick); returned at each appointment

Low – 29 % (12/42)

Medium: 5 complete data (all consecutive diaries), 7 partial data, ‘reasonable degree of consistency’ (SELF: interview) with occasional incorrect completion.

Unclear: ‘maybe it’s not a 100 % accurate’ (SELF: interview).

Case 5: SUPER

Single-blind feasibility trial comparing PFMT and lifestyle advice after surgery for pelvic organ prolapse to advice leaflet only (3 NHS PT centres, home)

Experimental (n = 28): Number of sets of PFMT undertaken per day (12 weeks)

▪Interview

▪Sample diary

▪Results publication [47]

▪Participant information sheet

1 A5 booklet (folded A4) (×5); 21 items per week; frequency (number), comments (open); returned at each appointment

Low – 29 % (8/28) ‘about half of those I think that returned all five diaries all completed’ (SUPER: interview).

Variable: ‘some of them are completed very well and very thoroughly and others are quite hard to understand really in what they actually meant’ (SUPER: interview).

Unclear: ‘some of them were … pretty accurate and pretty well filled in’ (SUPER: interview).

Case 6: EVIDEM-E

Single-blind pragmatic substantive trial comparing an individually tailored progressive walking regimen with exercise therapist support to usual care in patients with behavioural and psychological symptoms of dementia (Home/community in several inner city, urban and semirural locations)

All participants (n = 131 patient-carer dyads): number and duration of walks per day plus qualitative walk data. Intensity (rating of perceived exertion) for experimental group only (12 weeks)

▪Interview

▪Sample diary

▪Results publication [48]

▪Protocol publication [49]

▪Viewing of anonymised diary dataset

▪Participant information sheet

▪Trial website

49 page A5 booklet (×1); 64 items per week; frequency (yes/no, number), duration (number), intensity (Borg scale – circle response), comments (open), whether completed set course

(Yes/No), Why did not go out (circle response); collected at end of study

Medium: 68.7 % overall, intervention: 77.6 % (52/67), control: 59.4 % (38/64).

Variable: observed diary database and some well-completed, some little to no data.

Unclear: ‘I don’t think it’s reliable at all’ (EVIDEM-E: interview).

Case 7: TOMAS

Single-blind substantive trial (n = 568) comparing verbal advice, local travel information and PT/OT rehabilitation to verbal advice and travel leaflets only to improve stroke survivor outdoor mobility (Home/community in 15 NHS stroke services throughout Scotland, England and Wales)

All participants: number of journeys made per day (12 months)

▪Interview

▪Sample diary

▪Results publication [50]

▪Protocol publication [51]

▪Conference presentation relating to diaries [52]

▪Diary data coding sheet

12–13 page A5 booklet (×1); 14 items per week; frequency (number), falls (circle f); posted back monthly

Medium–high: 70.6 % of all expected travel diaries received and assigned. 89.4 % (508/568) returned at least one diary.

Medium: 55.1 % returned all diary pages for 12 months. ‘Most of them you could use the data from them’ (TOMAS: interview). Extensive variations in recording required development of a data entry coding sheet.

Unclear: ‘I imagine there was some people who didn’t keep them as accurate’ (TOMAS: interview).