Table 1.

Summary of Included Studies Characteristics (Author, Year, Setting, Study Design, Participants), Results (Medical Problems Concerned with Uncertainty, Definitions of Diagnostic Uncertainty, Linguistic Realisation, Communication Strategies, and Impact of Communication of Diagnostic Uncertainty to Patients) and Quality Appraisal (QATSSD) Score

Author and year	Location and study setting	Design/method	Number (n) of participants (% female (F))	Doctor age (yrs) and experience; Patient age (yrs)	Medical problems concerned with uncertainty (‘big three’: vascular, infections, and cancers18)	Definitions of diagnostic uncertainty/expressions of uncertainty	Communication strategy and linguistic realisation for diagnostic uncertainty	Impact of communication of diagnostic uncertainty on patient reaction and patient experience of care	QATSSD Score (%)
Quantitative studies
Bhise 201836	USA: convenience sample of parents living in a large US city	Quantitative: experimental vignette-based study design with surveys	Patients (parents) n=71 (89% F)	Patient (parent) mean age per intervention group 1. 36.3 yrs 2. 38.8 yrs 3. 41.4 yrs	• Abdominal pain (lasting 3 weeks)	Not given	Strategy: • Explicit disclosure • Implicit disclosure • Reasoning (differential diagnosis) Realisation: - Negated declarative (e.g. ‘not sure’) - modal verb (e.g. ‘could be caused by’) - Modal adverb/adjective (e.g. ‘most likely’) - Declarative (e.g. ‘could be caused by Disease A vs. Disease B or Disease C’)	Experience of care: • Explicit expression associated with less patient trust, adherence, and perceived doctor competence than implicit. • Reasoning favourable strategies (differentials) resulted in better patient experience of care.	73.8
Gerrity 199023	USA: doctors stratified by specialty (e.g. family medicine and internal medicine): half were licensed in North Carolina and half in Oregon.	Quantitative: questionnaire to doctors	Doctor n = 428 (12% F)	Doctor mean age: 46 ± 13 yrs Doctor mean experience: 20 ±14 yrs	No mention	Not given	Strategy: • Omission • Explicit disclosure Realisation: - Negated declarative (e.g. ‘I don’t know’)	Not mentioned	88.1
Gerrity 199237	USA: Doctors stratified by specialty (e.g. family medicine and internal medicine): half were licensed in North Carolina and half in Oregon.	Quantitative: questionnaire to doctors	Doctor n = 428 (12% F)	Doctor mean age: 46 ± 13 yrs Doctor mean experience: 20 ±14 yrs	• Abdominal pain • Sore throat • Chest pain (vascular)	Not given	Strategy: • Omission • Explicit disclosure Realisation: Not available	Patient reaction: Doctors withholding uncertainty because of potential to dissatisfy or confuse patient.	76.2
Ogden 200238	UK: Six general practices in the south-east of England	Quantitative: cross-sectional design, matched questionnaire	GPs n = 66 (42% F) Patients n = 550 (69% F)	GPs mean age: 44.86 ± 9.03 yrs Patient mean age: 48.09 ± 17.32 yrs	• No mention	Expression of uncertainty: ‘Expression of uncertainty […] were conceptualised as either behavioural expression (e.g. using a book or a computer or referring to hospital) or verbal expressions (e.g. “I don’t know” and “I’m not sure about this”)’ p. 172	Strategy: • Explicit disclosure • Implicit disclosure • PCC (reassurance) • Reasoning (information seeking) Realisations: - Negated declarative (e.g. ‘I don’t know’, ‘I’m not sure about this’) - Declarative (e.g. ‘I need to find out more’, ‘let’s see what happens’) - Modal verbs (e.g. ‘this might be..’) - Introductory phrase (e.g. ‘I think’)	Experience of care: • Explicit disclosure was associated with less patient confidence. • Patients rated verbal explicit statements worse than doctors. • Older patients with more experience with doctor tolerated uncertainty better. • Information seeking and implicit forms of expression of uncertainty seen as benign and even positive.	76.2
Olsen 201839	US: The University of Minnesota Paediatric Residency Program (commonly encountered clinical situations in general paediatrics)	Quantitative: Two-phase simulation-based educational intervention	Doctors (residents) n=23 Patients (simulated) n=4 (50% F)	Doctor experience: 2 yrs (Simulated) Patient age range: 0–6 yrs	• Hypotonia • Features consistent with Trisomy 21 • Ongoing fever (infection) • Viral infection (infection) • Lymphadenopathy • Malignancy, leukaemia (cancer) • Elevated inflammatory markers • Fatigue • Pancytopenia • Abnormal movements and ‘spells’	Diagnostic uncertainty: ‘The subjective and often appropriate perception that a clear and accurate explanation of a patient’s health problem is not able to be determined at this point in time.’ p.244 (modified from12)	Strategy: • Explicit disclosure • PCC (reassurance) • PCC (empathy) • Reasoning (diagnostic process) • Reasoning (eliminate/candidate diagnosis) Realisation: Not available	Not mentioned	83.3
Storten-beker 201940	NLD: GP clinic	Quantitative: analysis of video-recorded consultations, pre-post consult questionnaire for patients. Regression modelling	Doctors n=18 (N/A) Patients n =82 (60% F)	Doctor (GP) age not given Patient mean age: 52.6 (range 18–86)	• Chest pain due to acid reflux (Medically explained symptom) • Vague chest pain (MUS) (vascular)	Expression of uncertainty: ‘Frequent expressions of implicit uncertainty such as uncertain verbs (e.g. “could”, “I think”), lexical items (e.g. “probably”, “maybe”) and pragmatic particles (e.g. “sort of”)’ p. 2350	Strategy: • Implicit disclosure Realisation: - Modal verb (e.g. ‘could’) - Modal adverb/adjective (e.g. ‘maybe’, ‘probably’) - Adverb/adjective (e.g. ‘sort of’) - Introductory phrase (e.g. ‘I think’)	Patient reaction: • No relationship between implicit uncertainty and patient anxiety.	57.1
Mixed methods studies
Cousin 201341	SWI: Study 1. Vignette: analogue patients, ‘Doctor’ not specified Study 2: Interactions: GPs from the French-speaking part of Switzerland	Mixed: vignette-survey of patients and videotaped actual doctor-patient interactions and patients satisfaction survey	Vignette: patients n=120 (50% F) Interaction: doctors (GPs) n=36 (44% F) Patients n = 69 (47% F)	Vignette: patient mean age: 36.02 ± 12.51 yrs Interaction: doctor (GPs) mean age: 47.18 ± 9.55 yrs Patient mean age: 50.72 ± 18.19 yrs	• Back pain • Regular check-ups • Hypertension	Expressions of uncertainty: ‘Direct and indirect verbal expressions of uncertainty. Examples of direct expressions include “I don’t know” and “I have difficulty in answering this question”. Examples of indirect expressions include, for instance, certain adverbs (e.g. “probably”, “maybe”), probability statements (“There’s a good chance that…”), modal verbs (e.g. “might”, “may”, “should”) and conditional sentences (“If you feel better in a week…”)’ p. 927	Strategy: • Explicit disclosure • Implicit disclosure Realisation: - Negated declarative (e.g. ‘I cannot tell you’) - Modal adverb/adjective (e.g. ‘maybe’, ‘probably’) - Modal verb (e.g. ‘might’, ‘should’) - Conditional (e.g. ‘if you feel better in a week’) - Probability statement (e.g. ‘there’s a good chance that’)	Experience of care: Less patient satisfaction associated when female doctor communicated diagnostic uncertainty (no impact with male doctor). This effect only seen with male patients.	77.1
Epstein 200742	USA: family doctors and general internists in Greater Rochester area	Mixed: patient survey and audio recording of doctors interacting with a simulated patient	Doctors n=100 (23% F) Patients (survey) n=4746 (62% F)	Survey: patient mean age: 45 yrs (range 18–65) Simulated patients: presented as 48-year-old	• Chest pain (GERD role or medically unexplained symptoms (MUS) vascular, ?cancer) • Fatigue (MUS) • Dizziness (MUS) • Emotional distress (MUS)	Not given	Strategy: • Explicit disclosure • PCC (empathy) Realisation: - Negated declarative (e.g. ‘I don’t know’)	Experience of care: Explicit expression of uncertainty not associated with lower rating of doctor satisfaction/trust/autonomy/support/knowledge.	85.4
Gordon 200025	USA: university-affiliated general medicine clinic. General Medicine Clinic of the Portland, Oregon Veterans Affairs (VA) Medical Centre	Mixed: surveys of clinician response to uncertainty, patient behaviours and satisfaction with use of expression of uncertainty in consultations	Doctor n=43 (35% F) Patients n=43 (0% F)	Doctor age: not available Patient mean age: 62 yrs (range: 26–78) Doctor experience: 2 yrs (n=14), 3 yrs (n=14), staff doctors (n=15)	• Chronic illness requiring medication	Expressions of uncertainty: ‘[…] defined as a direct and unambiguous statement of uncertainty (for example, “I don’t know” or “It’s not clear”).’ p. 61	Strategy: • Explicit disclosure • PCC (empathy) • PCC (information giving) Reasoning (eliminate/candidate diagnosis) Realisation: - Negated declarative (e.g. ‘it’s not clear’) - Approximator (e.g. ‘pretty much normal’) - Introductory phrase (e.g. ‘my guess is’) - Adverb/adjective (e.g. ‘reportedly’) - Probability statement (e.g. ‘there’s a good chance that’)	Experience of care: Doctor explicit expressions of uncertainty were associated with greater patient satisfaction only when using positive talk, partnership building, and information giving. These patients wanted more information, and got more as a result.	58.3
Qualitative studies
Arborelius 199143	SWE: four health care centres	Qualitative: comments on videotaped consultations	Doctors n=9 (44% F) Patients n=14 (36% F)	Doctor mean age: 40 (range 35–50) Patient age: 20–97 yrs Doctor mean experience: 13 yrs (range 9–27)	• Weight loss • Loss of appetite • Pectoral and shoulder pains • Palpitation of the heart • Muscular rupture • Rheumatism • Hard life situation • Cold • UTI (infection) • Rectal pain	Not given	Strategy: • Omission Realisation: Not available	Not mentioned	33.3
Clarke 201444	UK: tertiary referral centre and GP clinics in Southern England	Qualitative: thematic analysis of qualitative semi-structured interviews	Doctors n=9 (N/A) Patients (parents) n=21 (86% F)	Not available	• New diagnosis of acute leukaemia (cancer)	Not given	Strategy: • Reassurance (safety netting) • Patient-centred communication (PCC, empathy) • Reasoning (eliminate/candidate diagnosis) Realisation: Not available	Not mentioned	83.3
Heath 199245	UK: general practice consultants gathered in various settings throughout the British Isles.	Qualitative: (No further methods stated)	Not given	Not given	• Ulcer • Anxiety • (Wear and tear) arthritis • Conjunctivitis (infection)	Not given	Strategy: • Explicit disclosure • Implicit disclosure • Embodied action • Reasoning (eliminate/candidate diagnosis) Realisation: - Negated declarative (e.g. ‘I wouldn’t know’) - Introductory phrase (e.g. ‘I think’) - Adverb/adjective (e.g. ‘not a totally typical’) - Interrogative (e.g. ‘if I was to say to you …?’) - Conditional (e.g. ‘I’m wondering if you’ve got…’) - Intentional vagueness (e.g. ‘you know’) - Hesitation	Patient reaction: • Patient does not respond, when doctor provides candidate diagnosis. • Patient responds with own opinion and lay perspective without challenging doctor when doctor presents diagnosis as tentative question. • Patient supports doctor’s diagnosis citing other sources (e.g. spouse).	19.1
Heritage 201946	USA: Western and Southern US clinical practices (2003–2005)	Qualitative: conversation analysis of video-recorded interactions + coding (grounded theory)	Doctor n=71 (N/A) Patients n= 212 (N/A)	Not given	• Upper respiratory symptoms • Musculoskeletal conditions	Expression of uncertainty: ‘[D]iagnoses delivered using epistemic modality, evidentialization, and epidemiologic generalization. […], mitigated diagnoses are just that: named medical conditions presented with some element of epistemic distancing.’ p. 267	Strategy: • Explicit disclosure • Implicit disclosure • Reasoning (eliminate/candidate diagnosis) Realisation: - Negated declarative (e.g. ‘I hope you don’t have…’) - Generalising declarative (e.g. ‘the most common reason for the lining to be irritated is…’) - Modal verb (e.g. ‘could’, ‘might’) - Perception verb (e.g. ‘looks like’, ‘it feels like’) - Modal adverb/adjective (e.g. ‘likely’, ‘maybe’, ‘probably’) - Impersonal pronouns (e.g. ‘it feels like’) - Introductory phrase (e.g. ‘what I think you have is..’) - Intentional vagueness (e.g. ‘you know’) - Gaze	Patient reaction: • Patient verbal responses are more extensive when mitigation present in diagnostic statements. • No direct gaze reduces patient likelihood to respond verbally.	42.9
Maynard 200347	USA: internal medicine clinic in hospital	Qualitative: case study ‘single case analysis’, conversation analytic research	Doctor n=1 (0% F) Patient n=1 (100% F)	Not given	• Mammograph result of lump (cancer)	Not given	Strategy: • Explicit disclosure • Implicit disclosure • Embodied action • PCC (interpersonal) • PCC (information giving) • Reasoning (diagnostic process) Realisation: - Negated declarative (e.g. ‘that’s not a hundred percent’, ‘[but we] can’t even tell: if yer having [X] or not’) - Modal verb (e.g. ‘this could be…’) - Perception verb (e.g. ‘it appears to be…’) - Impersonal pronouns (e.g. ‘it’s kind of like’) - Introductory phrase (e.g. ‘according to the ...’, ‘they see something that …’) - Intentional vagueness (e.g. ‘it’s kind of like…’) - Hesitation	Patient reaction: • Patient attempts to align understanding in response to intentional vagueness. • In response to hedging, patient is misaligned with doctor focus. • Humour with explicit disclosure allowed patient to accept uncertainty.	26.2
Maynard 200648	USA: Midwest university hospital primary care centre and Eastern US state primary care clinic	Qualitative: conversation-analytic investigations Video, case studies (1 good news, 1 bad news, 2 snippets)	Doctors n=3 (N/A) Patients n=3 (75% F)	Patient age range: 37–50 (patients 1 and 2) Patients 3 and 4 age not given	• Severe chest pain (vascular) • Leg pain • Armpit lump (cancer) • (Patient 2 has a definite cancer diagnosis)	Diagnostic uncertainty: ‘persistent medical complaints may go unexplained when a serious diagnostic possibility is excluded. This raises the specter of indeterminacy and uncertainty in clinical medicine. […clinicians] can be faced with symptoms of indeterminate origins and consequently must deal with uncertainty about a larger medical picture of the patient surrounding one particular episode of diagnostic news […]’ pp. 250, 276	Strategy: • PCC (reassurance) • Reasoning (eliminate/candidate diagnosis) Realisation: Not available	Patient reaction: • In response to serious diagnosis elimination patient tried to justify their visits because of ongoing symptoms.	23.8
Meyer 201952	USA: paediatric clinicians at two large academic medical institutions in Texas	Qualitative: semi-structured, face-to-face interviews	Doctors n=18 (65% F)	Not specified Doctor experience: 0–16 yrs	• Cough • Fever (?infection) • Headache • Vomiting • Abdominal pain	Diagnostic uncertainty: ‘Subjective perception of an inability to provide an accurate explanation of the patient’s health problem’ p. G108 (adopted from12)	Strategy: • Omission • Explicit disclosure • Implicit disclosure • PCC (reassurance) • PCC (empathy) • PCC (managing expectations) • PCC (information giving) • Reasoning (diagnostic process) • Reasoning (eliminate/candidate diagnosis) • Reasoning (information seeking) Realisations: - Negated declarative (e.g. ‘we don’t know what’s going on’)	Patient reaction: • Patients with lower education levels were less engaged with less request for details. Patients with higher education levels were more engagement but had more discomfort with uncertainty. • Patients from some cultural backgrounds [unspecified] regarded uncertain doctors less positively. • Fear, frustration, grief, anxiety in response to uncertainty. Empathy (listening) and planning was used to deal with these emotions. Experience of care: • Explicit honest expression of uncertainty led to more trust in doctor.	78.6
Paton 201749	UK: GP clinic	Qualitative: case study	Patient n = 1 (0% F)	3-year-old boy	• Wheeze and respiratory symptoms • Chest infection (infection)	Not given	Strategy: • PCC (reassurance) • PCC (empathy) • PCC (information giving) • Reasoning (diagnostic process) Realisation: Not available	Patient reaction: • Empathy, explaining the diagnostic process and tailored information giving were strategies used in response to patient parents’ frustration at uncertainty. Parents felt reassured.	23.8
Peräkylä 199850	FIN: four Finnish primary care health centres	Qualitative: conversation analysis of video-recorded interactions	Doctors n=14 (N/A) Patients n > 100 (N/A)	Not given	• Joint infection • Bacterial infection • Cartilage injury	Not given	Strategy: • Explicit disclosure • Implicit disclosure • Embodied action • PCC (reassurance) • PCC (information giving) • Reasoning (diagnostic process) Realisation: - Negated declarative (e.g. ‘but no bacterial infection seems to be there’) - Perception verb (e.g. ‘here appears to be…’, ‘seems to be …’) - Intentional vagueness (e.g. ‘things like that’) - Impersonal pronouns (e.g. ‘it really behaves so much as if..’) - Hesitation	Not mentioned	42.9
Peräkylä 200651	FIN: four Finnish primary care health centres	Qualitative: conversation analysis of video-recorded interactions	Doctors n=14 (N/A) Patients n > 100 (N/A)	Not given	• Joint infection • Bacterial infection	Not given	Strategy: • Implicit disclosure • PCC (information giving) • Reasoning (diagnostic process) • Reasoning (eliminate/candidate diagnosis) Realisation: - Declarative (e.g. ‘It’s probably a bit the…’) - Modal adverb/adjective (e.g. ‘probably’) - Perception verb (e.g. ‘the [X] feels …’, ‘seems to be …’) - Introductory phrase (e.g. ‘As tapping on the vertebrae didn’t cause any pain […] it suggests a …’) - Intentional vagueness (e.g. ‘a bit of …’) - Impersonal pronouns (e.g. ‘it suggests a …’)	Patient reaction: • Communication of uncertainty led to longer patient verbal responses (weak association). • Plain assertions led to passive reaction of patient in one case. • Explaining evidence as a way to manage diagnostic uncertainty when discrepancy exist between patient and doctor.	38.1

Abbreviations: SWE, Sweden; GERD, gastroesophageal reflux disease; GP, general practitioner; MUS, medically unexplained symptom(s); US, United States; UK, United Kingdom; PCC, patient-centred care; SWI, Switzerland; FIN, Finland, NLD, Netherlands