Table 1 Summary of targeted review documents relevant to proxy reporting

Regulatory or scientific agencies

FDA PRO guidance [3]

Yes

Discouraged, particularly for symptoms known only by the patient

Someone other than the patient

Proxy-patient perspective

Proxy reports as patient; observer reports on something observable and may also provide judgment

FDA patient-focused drug development (PFDD) guidance [12, 13]

Yes

Discouraged, may not be reflective of patient’s feelings; For patients who are unable to self-report, eliciting behaviors caregivers observe can help avoid proxy reporting

Caregiver; someone other than the patient or health professional; ObsRO are reported by parent, caregiver, or someone who observed the patient in daily life

Proxy-patient perspective

None; ObsROs can report on observations but do not include judgment

FDA PFDD guidance appendices [14]

Yes

Discouraged, but ObsRO recommended in instances where it is impossible to have self-report

Someone who is not the patient; ObsRO are reported by parent, caregiver, or someone who observed the patient in daily life

Proxy-patient perspective

Discouraged for concepts only known by the patient (e.g. symptoms)

EMA Appendix 2 (use of PROs in oncology) [2]

Yes

Should be avoided, unless it is the only means of obtaining lost information. Use when patients cannot contribute (e.g. cognitive impairment, severe ill health)

Caregivers (non-clinical)

Proxy-patient perspective

Discouraged; ObsROs can report on observations but do not include judgement

CMS measures management system supplement on PROs [15]

Yes

Measure developer needs to consider whether to allow them

–

–

–

AHRQ user’s guide on registries for evaluating patient outcomes [4]

Yes

If the only available measure, for patients not physically or cognitively able to provide direct assessment

Close family or caregivers

Proxy-patient perspective

–

Non regulatory

PCORI standards for PROs [16]

Yes

When patients are unable to self-report

Caregivers

–

Observable events or behaviors

OECD strengthening health system [17]

Yes

For non-response in certain disease groups, use proxy/observer

Caregivers

–

References DEMQOL and DEMQOL-Proxy

NQF PRO-PM 2013 [18]

Yes

–

–

–

–

NQF PRO-PM roadmap 2020 [19]

Yes

Patients with severe physical or cognitive impairments

Caregivers, family members, or other people

–

–

Development of a model for a National Dementia Registry [20]

Yes

In advanced dementia cases, suggest a proxy version of QOL-AD

Primary caregiver

–

References QOL-AD

Professional society

ISOQOL minimum standards [21]

No

–

–

–

–

ISOQOL’s User Guide to Implementing PROs in Clinical Practice [11]

Yes

Patients requiring assistance (incapacities, mentally or cognitively limited)

Significant others, caregivers, physicians

–

–

ISOQOL Dictionary [22]

No

Observers cannot report on how another person feels but only on what they are observed to do

–

–

–

ISPOR emerging good practices task force: conceptual foundation [23]

No

Discusses observer rater to contrast with clinical professionals for performing ObsRO (vs ClinRO) assessments

Caregiver, parent, spouse, companion (not clinician)

–

–

ISPOR emerging good practices task force: rare disease [24]

Yes

ObsRO measure must be based on observations and are not proxy measures

–

–

–

Montreal accord on PROs [25]

Yes

–

Proxy is a special kind of observer with a “shared experience” [with the patient] that enables them to “report on the outcome”

–

Symptoms, physical impairments, behavior, and function

Recommendations/guidelines

SPIRIT-PRO [26, 27]

Yes

Per EMA/FDA, discourage use of proxies unless they provide the only means of obtaining information that may be lost

Someone other than the trial participant

Proxy-patient perspective

–

CONSORT-PRO [28]

Yes

In some instances it may not be possible for the PRO to be completed directly by the patient

Caregiver, clinician

Proxy-patient perspective

–

MORECare consensus on outcome measurement in palliative care [29]

Yes

Deterioration or unstable symptoms with physical and/or mental debility

Clinicians or non-clinicians; Capture factors that may influence proxy raters

–

–

Standard measure sets

ICHOM Oncology [30]

No

–

–

–

–

ICHOM Stroke [31]

Yes

For patients unable to respond

Proxy, clinician, or abstracted from medical records

–

Proxy should make a judgement on observable domains only

COMET [32]

No

–

–

–

–

Measure developers

EORTC Measures Manual (cancer) [33]

Yes

Currently exploring the value of proxy assessment, which may be useful in patients with major cognitive problems

–

–

–

Guidelines for Assessing QoL in EORTC Clinical Trials [34]

Yes

When difficult or impossible for patients to rate their own QoL (e.g. cognitive impaired or terminally ill)

A family member (e.g. partner or parent) or “care taker” (e.g. physician or nurse)

–

–

FACIT [35]

Yes

Suggest a sequence: (1) self-report; (2) self-report with assistance from an objective person; (3) self-report with assistance from a familiar person; (4) proxy report

A familiar person (e.g. friend or family member)

–

–

PROMIS [36]

Yes

Patients unable to answer on their own, such as people in the early stage of dementia, cognitive or communication deficits, and severe disease burden

Someone else

–

–

Individual measures (for conditions with expected high use of proxy report)

QOLAS (dementia) [37]

Yes

Measure includes an assessment from caregivers

Caregivers

Proxy-proxy perspective

Physical, psychological, social/family, daily activity, cognitive

QOL-D (dementia) [38]

Yes

–

Members of nursing staff or family well acquainted with the patient

–

Positive and negative affect, restlessness, attachment with others, spontaneity and activity

DEMQOL and DEMQOL Proxy (dementia) [39, 40]

Yes

Measure includes an assessment from caregiver; In people with severe dementia, only DEMQOL-Proxy should be used due to high levels of missing data for DEMQOL

Family caregiver

Proxy-patient perspective

Daily activities, health and well-being, cognitive function, social relationships, self-concept (all subjective)

Cornell-Brown scale for quality of life (dementia) [41]

Yes

Measure includes an assessment from caregiver; Professional completes the scale after interviewing patient and caregiver

Caregiver with daily contact with the patient

–

Mood ratings, ideational disturbance, physical signs, cyclic functions

QOL-AD (Alzheimer’s, dementia) [42, 43]

Yes

Proxy perspective can be obtained for any participant, but those with MMSE ≤ 10 may be unable to answer

Actively involved caregiver who lived with patient or spent every day with them

Proxy-patient perspective

Perceived quality of life

ADRQL (Alzheimer’s) [44, 45]

Yes

Measure includes an assessment from caregiver

Close family members or professional caregivers

–

Social interaction, awareness of self, feelings and mood, enjoyment of activities, response to surroundings

MD Anderson Symptom Inventory user guide (cancer) [46]

No

–

–

–

–

EORTC QoL Questionnaires (cancer) [47]

No

–

–

–

–

Neuro-QoL (neurological disorders) [48]

Yes

Patients unable to answer on their own, such as people in the early stage of dementia, cognitive or communication deficits, and severe disease burden

Caregiver proxy responders; Someone else

Proxy-patient perspective

–

ASCQ-Me (sickle cell) [49]

No

–

–

–

–

Palliative Care Research Cooperative Measurement Tool Library [50]

Yes

–

Family caregiver

–

Refer to individual PRO user manuals for specific details

EQ-5D [51]

Yes

Proxy versions developed for cases where patients are mentally/physically incapable of self-reporting

A caregiver who knows the patient well (e.g. parent, physician, nurse)

Proxy version 1 is Proxy-proxy perspective; Proxy version 2 is Proxy-patient perspective

Mobility; self-care; usual activities; pain/discomfort; anxiety/depression; health today