Study | Research question | Geographical setting | Healthcare setting | Participants | Process or decision | Patient characteristics | Possible drivers of variation | Condition | Findings |
---|---|---|---|---|---|---|---|---|---|
Adams et al, 2014 [53] | Identification of mechanisms driving differential diagnoses and disparities that are common to black and white people in both countries; examination ofbetween-country variations due to cultural and health care system differences | UK and USA | Primary care | Physician | Diagnosis, referral, prescription | Ethnicity (black, white) | Information processing, patient cues, knowledge used, healthcare system. | Depression | There was little bias in doctors' decisions overall. UK doctors had greater clinical uncertainty in diagnosing depression amongst black than white patients . Doctors focused more on black patients' physical than psychological symptoms and more often tended to identify endocrine problems. |
Begeer et al, 2008 [54] | Whether ethnic background influences the likelihood of pediatricians’ references to Autism when using clinical judgments versus ratings of explicit diagnostic categories | Netherlands | Child health setting | Physician | Diagnosis | Ethnicity (Dutch vs Moroccan or Turkish) | Spontaneous vs prompted likelihood; physician characteristics | Autism | Spontaneous clinical judgements resulted in ethnic bias; this bias disappeared when doctors were prompted to consider autism. |
Bernardes et al, 2013 [55] | Whether physician sex moderates the effects of patient (distressed) pain behaviours and diagnostic evidence of pathology on treatment prescriptions and referrals; explore the mediating role of pain credibility judgments and psychological attributions on these effects | Portugal | Primary care | Physician | Referral, prescription, assessment | Gender | Physician sex, clinical cues (evidence of pathology, distress) | Chronic lower back pain | Confirming the hypothesis, physician sex moderates the influence of clinical cues on pain management practices: evidence of pathology had a larger effect on male than on female physicians’ referrals to psychology/psychiatry. |
Bories et al, 2018 [56] | To test the hypothesis that physician uncertainty aversion impacts medical decision making for older patients with acute myeloid leukaemia | France | Acute | Physician | Prescription | Age (note clinically relevant) | Physician demographic, occupational, behavioural characteristics | Acute myeloid leukaemia | Physician attitudes to risk influenced chemotherapy decisions for older patients. Physicians opting for intensive chemotherapy (IC) had higher aversion to uncertainty and treated fewer patients annually, than the low IC group but were similar in age, hierarchical status or years of experience. |
Burgess et al, 2014 [57] | To test the hypothesis that racial biases in opioid prescribing would be more likely under high levels of cognitive load | USA | Primary care | Physician | Prescription | Ethnicity (black, white) | Physician cognitive load | Chronic low back pain | Hypotheses were partially confirmed. Cognitive load altered ethnic inequalities in prescribing patterns in different ways for male and female physicians. Under high cognitive load, male physicians were more likely to prescribe opioids for White patients; while under low cognitive load, they were more likely to prescribe opioids for Black patients. Female physicians’ bias toward prescribing opioids to Black patients was stronger under greater cognitive load. |
Burt et al, 2016 [58] | To examine whether South Asian people rate GP consultations similarly to White British people, in order to understand why minority ethnic groups often give poorer evaluations of primary care | England | Primary care | Public | Consultation style | Ethnicity (South Asian, white) | Patients' ratings of quality | Persistent cough, perforated ear drum, painful elbow generalised numbness | Respondents from a Pakistani background rated communication in simulated GP consultations significantly more positively than their White British counterparts (contrary to the hypothesis that South Asians’ poorer evaluations of primary care experience is due to higher expectations of care). |
Daugherty et al, 2017 [59] | To test the hypotheses that physician gender bias would have little effect on treatment decisions for the male patient and would result in lower use of cardiovascular tests among gender-biased physicians for female patients | USA | Â | Physician | Diagnosis | Gender | Implicit bias | Coronary artery disease | Hypotheses were partially confirmed; cardiologists who associated risk taking more with men than with women were more likely to view angiography as useful to diagnose male versus female patients but equally likely to recommend stress testing. Physicians were less certain of diagnosis in women than men. |
Elliott et al, 2016 [60] | To test whether hospital-based physicians use different verbal and/or nonverbal communication with black and white simulated patients and their surrogates. | USA | Acute | Physician | Consultation style | Ethnicity (black, white) | Verbal and non-verbal communication between patient & physician | Metastatic gastric and pancreatic cancer | Physicians used similar verbal but different nonverbal communication behaviours with black and white patients. |
Fischer et al, 2017 [61] | To test whether patient requests for specific opioid pain medication would lead physicians to classify them as drug-seeking and change management decisions | USA | Primary care | Physician | Prescription | Ethnicity (black, white) | Patient (drug seeking) behaviour | Pain (sciatica) | Physician suspicion of drug-seeking behaviour was much higher when patients requested opioid medication. Physician suspicion of drug-seeking behaviour did not vary by patient characteristics, including gender and race. |
Gao et al, 2019 [62] | To test whether Chinese favour family-centred decision making while European Americans favour shared decision making in depression care | USA | Other - mental health | Public | Mode of decision making - hospital or community care | Race, nationality | Acculturation, preferences for care | Depression | Hypotheses were confirmed; Chinese preferred family-centred decision making while Americans preferred shared decision making. Chinese living in America paralleled European Americans. |
Green et al, 2007 [63] | To test whether implicit or explicit race biases predict physicians' decisions to give thrombolysis for acute myocardial infarction | USA | Acute and primary care | Physician | Diagnosis, prescription | Ethnicity (black, white) | Physician implicit bias | Acute myocardial infarction | Hypothesis was confirmed. As physicians’ pro-white implicit bias increased, so did their likelihood of treating white patients and not treating black patients with thrombolysis. |
Hirsh et al, 2009 [64] | To test whether gendered expectations of pain and facial pain expressions influenced pain assessment and treatment disparities in nurses | USA | Acute | Nurse | Prescription | Age, sex, race | Gender role expectations of pain (sensitivity, endurance, willingness to report), high/low pain facial expression | Pain appendectomy | Hypotheses were partially confirmed; nurses’ gender role expectations of pain didn’t influence decisions but pain expression did. Nurses generally rated female, African American, older patients’ pain higher and were more ready to prescribe opioids. |
To test whether patient-provider racial concordance and patient ethnic salience is associated with 1) provider pain assessment 2) attitudes toward referral for traditional healing practices for indigenous patients | USA | Primary care | Other clinical professional | 1) Prescription 2) referral | Ethnicity (Indigenous American - high/low ethnic salience) | Racial concordance (patient & physician) | Chronic lower back pain | 1) Indigenous providers rated patient with higher Indigenous ethnic salience more congruently with the self-reported pain ratings 2) Provider–patient racial concordance increased likelihood of consulting with and referring patients to traditional healing practices. | |
1) Whether physician certainty is associated with decision making. Explore variations, by health care system, patient characteristics 2) whether observed disparities in CHD decision making are influenced by priming physicians to consider CHD. | USA, Germany and England | Primary care | Physician | Diagnosis, referral, prescription, lifestyle recommendations | Age, gender, ethnicity (black, white), SES | Diagnostic certainty, healthcare system, physician priming | CHD | 1) Certainty was positively correlated with test ordering, prescriptions and specialist referrals. Physicians were least certain of CHD diagnoses when patients were younger and female. 2) Primed physicians were more likely to order CHD-related tests and prescriptions than unprimed but main effects for patient, gender and age remained. | |
Mckinlay et al, 2012 [68] | Whether physicians’ decisions to diagnose diabetes vary by race/ethnicity (after controlling for SES, age, and gender). | USA | Primary care | Physician | Diagnosis | Age, gender, ethnicity (black, Hispanic, white), SES | Effects of SES on ethnicity | Diabetes | Primary care physicians’ vignette diagnosis was patterned by race/ethnicity (rather than by SES). [Undiagnosed signs of T2DM in the community was patterned by SES rather than race/ethnicity.] |
Papaleontiou, et al 2017 [69] | Understanding why older thyroid cancer patients are not being referred to high-volume surgeons. | USA | Primary care | Physician | Referral | Age | Physician training, patient volume, discipline & patient preferences | Cancer | Endocrinologists and physicians treating more than 10 thyroid cancer patients per year were more likely to refer older thyroid cancer patients than primary care physicians. Patient preference, transportation barriers and confidence in local surgeon were commonly reported reasons to decrease likelihood of referral. |
Samuelsson et al, 2014 [70] | Disentangle a number of determinants on addiction care practitioners' perceptions of the severity of alcohol and drug consumption in clients. | Sweden | Addiction | Other | Referral (eligibility for services), perceptions of severity | Age, gender, ethnicity, SES, family circumstances | % variance due to vignette, professional and work unit | Substance use | Practitioners of different professional backgrounds and workplaces judge alcohol and drug consumption by different norms, and this was also influenced by characteristics of the users. |
Shapiro, et al. 2018 [71] | Whether neonatologists show implicit racial and/or socioeconomic biases and whether these are predictive of recommendations at extreme periviability | USA | Acute | Physician | Care: comfort vs intensive (e.g. resuscitation) | Race, SES | Implicit bias | Periviability | Hypotheses were in part confirmed. Physicians with implicit socioeconomic bias were more likely to recommend comfort care to high than low SES vignettes but did not appear influenced by implicit racial bias. |
Sheringham et al, 2017 [72] | How patients' clinical and sociodemographic characteristics influence GPs’ decisions to initiate lung cancer investigations | England | Primary care | Physician | Diagnosis | Age, gender, ethnicity (black, South Asian, white), SES | Information elicited, physician attributes | Respiratory symptoms | The information GPs elicited from patient vignettes influenced their decisions but did not explain observed ethnic inequalities in cancer investigations |
Tinkler et al, 2018 [73] | Whether appointment offers to new US primary care patients who mention concerns about smoking or weight differ from offers to patients with no health concerns (healthy patients) | USA | Primary care | Other | Appointment offer | Insurance status, race/ethnicity, and gender | Health concerns (smoking/weight concerns vs healthy); state-level Medicaid expansion status | Disease prevention | Patients with smoking concerns were no more likely to be offered new patient appointments than healthy patients and less likely than those with weight concerns. Insurance status influenced access. |
Wiltshire,et al. 2018 [74] | Whether concordance leads to higher ratings of trust in physicians amongst African American women race, gender and age | USA | Primary care | Public | Trust | Race - age, gender | Concordance | Breast exam | Older African-American women did not rate race, gender or age-concordant doctors higher on trust; instead they rated white, older male higher on competence than African-American older females. |