Table 1.
Details of included papers identified during search 1988–2020
| Authors and title | Aims/objectives | Country of study | Article type, research methods, data collection analysis and limitations | Key results |
|---|---|---|---|---|
| Qualitative research | ||||
| Physician responses to ambiguous patient symptoms. Seaburn et al. (22) |
To examine how primary care physicians respond to ambiguity. | USA | Type: primary research—qualitative Methods: observational study of n = 23 GPs using standardized patients. Analysis: thematic analysis. Limitations: a relatively small sample of clinicians, utilizing qualitative analysis. Many not be generalizable to wider population of GPs. Use of standardized, as opposed to real, patients. |
• Physicians’ responses to ambiguous symptoms were categorized into two primary patterns: high partnering (HP) (43%) and usual care (UC) (57%). |
| • HP was characterized by greater responsiveness to patients’ expression of concern, greater acknowledgment of symptom ambiguity and solicitation of patients’ perspectives on their problems. | ||||
| • UC was characterized by denial of ambiguity and less inclusion of patients’ perspectives on their symptoms. | ||||
| • When confronted with ambiguous symptoms, study physicians almost uniformly either denied ambiguity or coupled acknowledgment of ambiguity with a directive. | ||||
| Patients’ experiences of cancer diagnosis as a result of an emergency presentation: a qualitative study. Black et al. (26) |
To capture in detail the experiences of patients whose cancer was diagnosed following an ED visit to understand how emergency presentations arise and identify where there is scope to improve outcomes. | UK | Type: primary research—qualitative Methods: interviews with n = 27 patients to assess their experiences of health care services before diagnosis. Analysis: thematic analysis of interviews. Limitations: a limited number of potential participants were approached for interview. Results may not be generalizable to non-cancer patients. |
• Patient concerns about discovering they have cancer, and about the harms or embarrassment of investigations, may have acted as particular incentives to adhering strongly to a benign diagnosis once given. |
| Challenges and strategies for general practitioners diagnosing serious infections in older adults. McKelvie et al. (25) |
To explore issues GP face in the diagnosis of serious infection in adults. | UK | Type: primary research—qualitative Methods: semi-structured interviews n = 28 GPs. Analysis: framework analysis of interviews. Limitations: self-selection bias in recruitment, as well as hindsight bias in GP accounts of cases. The study is specific to older patients, therefore may not be generalizable to general population. |
• Three main themes emerged: challenges leading to diagnostic uncertainty (patient complexity, atypical presentation, lack of knowledge about the patient); approaches to the patient (history-taking, physical examination and use of scoring systems), and management of diagnostic uncertainty (ordering investigations, and safety-netting). |
| • Investigations were sometimes used to resolve diagnostic uncertainty, but availability and speed of result limited their practical use. | ||||
| • Clear safety-net plans shared with patients and their families helped GPs manage ongoing uncertainty—GPs both discussed safety-netting plans with a third party if required, and gave specific conditions when they would expect the patient to return. | ||||
| • Shared responsibility by safety-netting gave GPs less anxiety. | ||||
| ‘Chasing a Ghost’: factors that influence primary care physicians to follow up on incidental imaging findings. Zafar et al. (27) |
To explore provider and patient characteristics that influence how primary care providers communicate and manage incidental image findings. | USA | Type: primary research—qualitative Methods: semi-structured interviews n = 30 primary care providers. Analysis: framework analysis of interviews. Limitations: this study was conducted at a single academic medical centre, which may limit the generalizability of these results. There was no patient involvement, so only able to give information on physician perspectives on decision-making. |
• In order to eliminate uncertainty, some primary care providers felt compelled but frustrated to pursue costly follow-up for incidental imaging findings of limited clinical importance. Others did not act on findings that were unfamiliar or occurred in an unusual clinical context when follow-up recommendations were not given. |
| • Some reported using a uniform approach to communicate and manage incidental findings, while others used pre-test probability to inform them on the significance of the finding, or adapted their approach according to their patient. | ||||
| • There was evidence of variation in how much information about the possible diagnosis causing incidental imaging findings primary care providers disclose to different patients. | ||||
| “I got my diagnosis on a yellow post-it note”: young adult cancer patients’ experiences of the process of being diagnosed with cancer. Hauken et al. (29) |
To explore how young cancer survivors experience the process of being diagnosed with cancer. | Norway | Type: primary research—qualitative Methods: a qualitative method founded on a phenomenological-hermeneutical approach was used, and included in-depth interviews with 20 young adult survivors (aged 24–35 years) with different cancer diagnoses. Analysis: thematic analysis of interview data. Limitations: small sample size. Findings cannot be generalized to the broader population. The study also had an underrepresentation of men and of the youngest age group (18–23 years). |
• The participants’ experiences of the diagnosis process were elaborated according to three main themes: (i) ‘I felt something was wrong, but…’ (ii) ‘The traumatic uncertainty’ and (iii) ‘The day my world collapsed’. |
| • ‘The traumatic uncertainty’ was further divided into two subthemes: (i) ‘But no one would tell’ and (ii) ‘To live in suspense’. | ||||
| • A common feature of this period was that the participants experienced a dearth of information and considerable worry and uncertainty; many did not fully understand the reasons for invasive and sometimes painful investigations. | ||||
| • Several participants had long periods awaiting investigation results, characterized by considerable anxiety. | ||||
| Quality improvements of safety-netting guidelines for cancer in UK primary care: insights from a qualitative interview study of GPs. Tompson et al. (18) |
To seek the insight of frontline GPs regarding proposed safety-netting guidelines for suspected cancer in UK primary care. | UK | Type: primary research—qualitative Methods: semi-structured interviews with GPs (n = 25). Interviewees were asked about their views and experiences of safety-netting and then presented with the safety-netting recommendations. Analysis: thematic analysis of interview data. Limitations: small sample of GPs, all from one region. 1/3 of the interviewees were personal contacts of the research team. |
• GPs were supportive of initiatives to optimize safety-netting. |
| • Sharing diagnostic uncertainty was thought to be helpful, but was only partially implemented. | ||||
| • GPs agreed that they should explain the uncertainty of the working diagnosis to patients, but suggested that their ability to do so was limited both by the length of consultations and by some patients preferring ‘black and white answers’. | ||||
| • Neither informing patients of all (including negative) test results nor ensuring recurrent unexplained symptoms are always flagged and referred were considered feasible. | ||||
| Communicating cancer risk in the primary care consultation when using a cancer risk assessment tool. Akanuwe et al. (30) |
To explore the perspectives of service users and primary care practitioners on communicating cancer risk information to patients, when using QCancer, a cancer risk assessment tool. | UK | Type: primary research—qualitative Methods: semi-structured interviews and focus groups with 36 participants (19 service users) and 17 primary care practitioners. Of the 19 service users, two had a previous diagnosis of cancer, and the rest had relatives or friends who had a previous diagnosis of cancer. Analysis: framework method of thematic analysis. Limitations: relative lack of diversity in the sample: all 19 service user participants were of White British ethnicity. Users all had personal experience of cancer, and their views might systematically differ from those without experience of a cancer diagnosis. |
• This paper discusses the communication of uncertainty regarding potential cancer diagnosis, in the context of patients presenting with undifferentiated symptoms. |
| • Participants suggested ways to improve communication of cancer risk information: personalizing risk information; involving patients in use of the tool; sharing risk information openly; and providing sufficient time. | ||||
| • A key theme was being open and honest: participants emphasized the importance of telling patients the truth about their health information, even if this led to increased worry. | ||||
| • Not being honest about the risk of cancer could affect trust between patients and practitioners. | ||||
| Improving the quality of care and patient experience of care during the diagnosis of lupus: a qualitative study of primary care. Amsden et al. (28) |
To better understand diagnostic delay and doctor–patient communication during the diagnosis of systemic lupus erythematous in patients without malar rash. | USA | Type: primary research—qualitative Methods: interviews with n = 8 primary care physicians. Analysis: thematic analysis. Limitations: small number of participants, self-selection bias, hindsight bias in recounting cases. |
• There were five domains related to diagnosis: initial assessment and tests, initial diagnosis and empiric treatment, timeliness of diagnosis, communicating with the patient and opportunities for improvement. |
| • In general, the physicians stressed the importance of being honest with the patient. Some would explain the approach to the diagnosis and outline the differential diagnosis. | ||||
| • The paper discusses the need to clearly communicate when a diagnosis is not yet final and when an ordered treatment is empiric in order to reduce diagnostic errors and delays. | ||||
| • Doctor–patient communication is critical to help the physician make sense of the symptoms, maintain trust and assure the patient that he or she is receiving appropriate care. | ||||
| Acute low back pain management in general practice: uncertainty and conflicting certainties. Darlow et al. (24) |
To explore GP beliefs about acute low back pain, and how these influence their clinical management. | NZ | Type: primary research—qualitative Methods: semi-structured interviews with n = 11 GPs. Analysis: thematic analysis. Limitations: small number of participants, centred around one presenting complaint. |
•Specific diagnosis was often not seen as achievable or helpful in the diagnosis of low back pain. |
| •GPs perceived patients as intolerant of uncertainty and wanting a diagnosis. | ||||
| •GPs either (i) gave a specific diagnosis in order to reassure patients, or (ii) were more open about the uncertainty. | ||||
| •Some participants provided diagnostic labels despite an inability to accurately diagnose, and against guideline recommendations. They felt simple musculoskeletal diagnoses were helpful to reassure (themselves as much as their patients) that pathology was not present, and to encourage exercise, which was therapeutic. | ||||
| Quantitative research | ||||
| Doctors expressions of uncertainty and patient confidence. Ogden et al. (16) |
To examine the impact of the way in which uncertainty was expressed (behaviourally versus verbally) on doctor’s and patient’s beliefs about patient confidence. Also examined the role of the patient’s personal characteristics and knowledge of their doctor as a means to address the broader context. | UK | Type: primary research—quantitative Methods: matched questionnaires sent to GPs and patients. Fourteen expressions of uncertainty were described, conceptualized as either behavioural (n = 8) or verbal (n = 6). Participants were asked about what effect they thought the expressions of uncertainty would have on patient confidence; they were asked to rate each of the expressions of uncertainty on a 5-point Likert scale ranging from `not at all confident’ (1) to `totally confident’ (5). Patients also asked how often they had been to see the GP in the past year, and how well they felt that they knew their GP rated on a 5-point Likert scale. Analysis: statistical analysis on GPs’ and patients’ ratings of the expressions of uncertainty. The data were also analysed using multiple regression analysis to explore the role of the patient’s personal characteristics and experience of their doctor in predicting their reaction to verbal and behavioural expressions of uncertainty. Limitations: use of analogue patients. Survey data asking participants to speculate on effects, rather than collecting actual effects of communicating uncertainty in real interactions. |
• Verbal expressions of uncertainty such as ‘Let’s see what happens’ were viewed as the most potentially damaging to patient confidence, and both GPs and patients believed that asking a nurse for advice would have a detrimental effect. |
| • In contrast, behaviours such as using a book or computer were seen as benign or even beneficial. | ||||
| • GPs and patients agreed about behavioural expressions of uncertainty, but the patients rated the verbal expressions as more detrimental to their confidence than anticipated by the doctors. | ||||
| • Patients who indicated that expressions of uncertainty would have the most detrimental impact upon their confidence were younger, lower class and had known their GP for less time. | ||||
| Mixed methods research | ||||
| Communicating and dealing with uncertainty in general practice: the association with neuroticism. Schneider et al. (17) |
To estimate the association with personality traits on handling of uncertain situations in general practice. | Germany | Type: primary research—mixed methods Methods: mixed methods approach: Qualitative analysis of focus group discussion. Interviews using cognitive think aloud technique. (10 GPs). Cross-sectional survey of GPs (228 responses). Analysis: statistical analysis (mean, standard deviation, ANOVA, Pearson correlation) of data set. Limitations: questionnaires developed with GPs who attend conferences and teach—bias towards those who these interests/critical thinking. |
• Neuroticism was positively associated with all PRU scales ‘anxiety due to uncertainty’, ‘concerns about bad outcomes’, ‘reluctance to disclose uncertainty to patients’ and ‘reluctance to disclose mistakes to physicians’. |
| • Neuroticism was negatively associated with the CoDU scale ‘communicating uncertainty’. | ||||
| • ‘Extraversion’, ‘agreeableness’, ‘conscientiousness’ and ‘openness to experience’ were significantly positively associated with ‘communicating uncertainty’. | ||||
| Strategies for managing uncertainty and complexity. Hewson et al. (20) |
To identify strategies involved in diagnosis and treatment of primary care problems that are uncertain and complex. | USA | Type: primary research—mixed methods Methods: (1)Observed videos (n = 10) of primary care physicians interacting with four standardized patients with complex/uncertain primary care problems, performed qualitative analysis to identify ‘strategies’. These strategies then rated by physicians in GIM (n = 19) on a 1–10 Likert scale for perceived importance in their primary care practice. (2)Transcripts of audiotapes of physician-standardized patient encounters analysed to record incidence of the strategies. Limitations: small numbers, identification of strategies for dealing with complexity (as well as uncertainty). |
• Nine strategies were identified, and each was rated as important to primary care clinical practice. |
| • Strategies include: eliminates alternative diagnoses by dealing with patient fears, giving reasons in the context of the patient’s belief system; and keeps diagnostic options open by making provisional diagnoses while keeping alternatives in mind. | ||||
| ‘Could this be something serious?’ Reassurance, uncertainty, and empathy in response to patients’ expressions of worry. Epstein et al. (21) |
To describe physicians’ responses to patients’ worries, how their responses varied according to clinical context, and associations between their responses and patients’ ratings of interpersonal aspects of care. | USA | Type: primary research—mixed methods Methods: multimethod study. For each physician covertly audio recorded two unannounced standardized patient (SP) visits. SPs expressed worry about ‘something serious’ in two scenarios: straightforward gastroesophageal reflux or poorly characterized chest pain with MUS. Also patient surveys (n = 50) given to real patients in the waiting room for each physician, measuring interpersonal aspects of care (trust, physician knowledge of the patient, satisfaction and patient activation). Analysis: qualitative coding of these transcripts, followed by descriptive, multivariate and lag-sequential analyses. Limitations: use of SPs, only two clinical scenarios. Expressions of uncertainty ‘I don’t know’, as opposed to acknowledgement of diagnostic uncertainty. |
• Biomedical inquiry and explanations, action, non-specific acknowledgment and reassurance were common, whereas empathy, expressions of uncertainty and exploration of psychosocial factors and emotions were uncommon. |
| • Expressions of uncertainty (e.g. ‘I don’t know’) were the least frequently observed response. | ||||
| • Physicians’ expressions of uncertainty (‘I don’t know…’) were not associated with lower patient ratings of the physician. | ||||
| Dealing with uncertainty in general practice: an essential skill for the general practitioner. O’Riordan et al. (2) |
To discuss the importance of managing uncertainty in primary care and to propose educational interventions which can improve this. | European | Type: primary research—mixed methods Methods: mixed methods: literature review and expert focus group consensus. Limitations: details of methodology for focus group consensus not included in manuscript. |
• Describes the management of uncertainty as an essential skill which should be included in educational programmes for both trainee and established GPs. |
| • The literature on dealing with uncertainty focuses largely on identifying relevant evidence and decision-making. | ||||
| • Highlights that there is a need for GPs to accept some uncertainty in complex situations to avoid burnout. | ||||
| • Suggests that a good doctor–patient relationship is vital, creating trust and mutual respect, developed over time with good communication skills. | ||||
| • Recommends sharing the degree of uncertainty with the patient. Evidence-based medicine should be used, including discussion of probabilities where available. | ||||
| • Notes that SDM encourages sharing uncertainty with patients, but the effect of doing this on patients has not been thoroughly researched. | ||||
| Identifying transparency in physician communication Robins et al. (19) |
To categorize physician communication, to gain understanding of what patients want to know and skill in conveying that information. | USA | Type: primary research—mixed methods Methods: analysis and coding of 263 audiotaped interactions between 33 primary care physicians and their patients in eight community-based, primary care clinics. Analysis: qualitative examination of audiotapes, with some quantitative analysis (e.g. the proportion of time spent on different behaviours in the consultation). Limitations: wide variation in transparency utterances across physicians. Study was not able to determine whether study patients noticed or valued the types of transparency utterances identified or what these utterances might have contributed to patients’ perceptions of their physicians. |
• Physicians proactively used five types of process transparency to preview speech and actions. Four types of content transparency were used to explicate diagnosis and treatment, demystify medical language and concepts and interpret biomedical information. Physicians spent the greatest proportion of clinic time explicating medical content. |
| • The typology ‘diagnosis rationale’ (Sharing personal thinking about patient’s diagnosis, especially what symptoms or signs are suggestive of a condition or distinguish one possible condition from another) was one of those that physicians engaged least with as a proportion of clinic time (<1.0% mean encounter time). | ||||
| The wrong diagnosis: Identifying causes of potentially adverse events in general practice using incident monitoring. Bhasale et al. (23) |
To identify how diagnostic incidents occur and to understand their preventable causes. | Australia | Type: primary research—mixed methods Methods: GPs anonymously reported incidents of potential harm using free text or structured responses. In this paper, n = 142 diagnostic incidents were examined. Analysis: descriptive statistics of quantitative data, and thematic analysis of qualitative data. Limitations: hindsight bias, as results rely on the accuracy of GPs reporting incidents. |
• Diagnostic incidents are often due to errors in judgement in the formation of a diagnosis Type A errors (inappropriately rejecting the correct diagnosis) and Type B errors (inappropriate hypothesis formation). Poor communication (both between health care professionals and between doctors and patients) often frequently contributes. |
| • Ineffective communication to the patient about the need for further assessment/poor understanding about the diagnostic process was the main factor in five incidents. | ||||
| • Patients did not always know what was expected of them, e.g. a case of delayed appendicitis diagnosis, in which the patient was not adequately informed about the provisional nature of the initial working diagnosis of gastroenteritis. | ||||
| • In another case, a patient did not return for planned X-rays, having not understood risk of unhealed scaphoid fracture. | ||||
| • Improved communication between patient and doctor (by being clear about the diagnostic uncertainty) could reduce diagnostic incidents. |