Abstract
Background:
Understanding patients’ experiences with diagnosis is crucial for improving care. Clarifying how best to ask about these experiences is an essential step in doing so.
Objective:
To elicit and then categorize diagnostic problems and mistakes in a national, population-based survey.
Methods:
Drawing from a nationally representative panel, respondents were asked whether they had experienced (directly or as a care partner) a diagnostic problem or mistake in the last 4 years. They then classified the experience as a mistake, a problem only, or a combination of both. Respondents who experienced multiple problems and mistakes reported on their most memorable experience. We compared responses among those reporting a problem, a mistake, or a combination of both.
Results:
A third [1216/3684 (33.0%)] of screened households reported diagnostic problems or mistakes involving themselves [697/3684 (18.9%)] or someone close to them [519/3684 (14.1%)]. A plurality [448/3684 (12.2%)] categorized these as a mistake, 382 (10.4%) as a problem, and 371 (10.1%) as both. Experiences reported as problems versus mistakes were equally likely to be associated with harmful consequences and concrete clinician responses. The distribution of problems and mistakes across the 3 categories did not significantly differ when reported by patients versus care partners, though there were checkered differences in the reported impact.
Conclusions:
Whether labeled as problems or mistakes, diagnostic experiences reported by patients and care partners are accompanied by substantial emotional, physical, and financial impacts. Responding to this full range of patient experiences is important for guiding improvements in diagnostic quality within learning health systems.
Keywords: Diagnostic errors, patient-reported outcomes, caregivers, surveys and questionnaires, national survey, health care quality
Since the turn of the century, responding to patients’ health care experiences has been a central tenet for improving health system performance.1,2 Doing so requires systematic means for collecting patient-reported experiences and outcomes, learning from that feedback, and incorporating insights into health policies and practices.3,4 Efforts to promote learning health systems increasingly incorporate pathways of influence for patients and care partners.5,6
The patient safety field has been slower to embrace patient feedback as a catalyst for improvement.7 Some skeptics saw safety events as too technical for patients to reliably assess.8,9 Others viewed key distinctions in safety discourse as challenging for patients to interpret. Safety events are conventionally grouped into those that are preventable (“medical errors”) and those that are not (“adverse events”). Each type is equally common in clinical settings,10 but safety-promoting interventions typically prioritize error identification, which patients or care partners have difficulty distinguishing from other adverse events.11–13
Despite these cautions, the patient safety field is turning to patient experience to better identify safety events and illuminate actionable forms of remediation.14,15 This includes efforts to understand and improve diagnostic quality,16,17 although arguably patients’ diagnostic experiences remain underutilized by learning health systems and quality improvement efforts.18,19 One persisting impediment involves a lack of clarity about how to describe the diagnostic events about which survey respondents are asked. Terminology varies in recent studies, ranging from asking about “medical errors” that involved diagnosis, to “adverse events” during diagnosis, to “patient safety concerns” or “breakdowns” in the diagnostic process.14,16,18,19 Appropriate labeling can be crucial, because asking respondents to make difficult attributions or to cast a pejorative light on trusted clinicians may induce underreporting of correctable events.15
This study uses an alternative approach to eliciting patient feedback, based on a 2-part label for difficulties during diagnosis. By asking about “diagnostic problems and mistakes” (abbreviated “diagnostic P&Ms”), we aim to reduce the cognitive and emotional burdens of assessing whether negative experiences involved “errors.” The dual terminology also encourages patients to report difficulties without understanding rigid or technical definitions. An inclusive “diagnostic P&Ms” also aligns with the growing interest in the patient safety field to promote diagnostic excellence rather than just correct failures, thereby enhancing timeliness, effectiveness, equity, and patient-centeredness.20
Fostering more inclusive reporting of diagnostic difficulties is essential to identify the full array of insights from which health systems and policymakers can learn.21,22 We report here on results exploring patients’ experiences using this dual-term approach in a nationally representative, population-based survey. We examine differences between respondents’ reports of diagnostic problems, mistakes, or events that combine both with respect to the impact on patients, perceptions of clinician involvement, and clinician responses to a diagnostic P&M.
CONCEPTUAL FOUNDATIONS
Research on how people make attributions regarding the causes and consequences of complex events dates back more than 50 years, with applications to patient expectations and behavior for about half as long.23,24 Applications to patient safety are more recent; their key insights regarding patient behaviors and voice are consistent with these longer-established models regarding blame and responsibility.12,13,15,25,26
First, when causal processes are complex, people hesitate to assign blame unless they are certain about identifying the culpable parties. Asking patients or care partners about experiences involving “errors” sets a high bar for causal attribution that discourages reporting of even quite harmful events.27 Second, the hierarchy in health care can make patients reluctant to suggest a clinician did something wrong; some patients feel daunted questioning clinical expertise, and are thus discouraged from reporting.24,28 Third, responses that require a sophisticated assessment of causes inhibit patient voice most for those least experienced in health care settings, since they tend to be less confident in judging if harms could have been avoided.24,29
These considerations make it unwise to frame patient surveys exclusively in terms of diagnostic errors. Offering multiple framing terms encourages responses from those less comfortable reporting their experiences. Finally, those whose diagnostic experiences are more complex do not need to sort out multiple causes to determine whether to participate in the survey.
METHODS
Survey Development
We developed a survey to elicit patient-centered insights and perceptions about diagnostic P&Ms. The survey included a series of closed-ended questions, informed by literature reviews and adapted from prior work on patient safety concerns (eg, medication errors).18,19,30,31 The survey also incorporated 10 open-ended questions to collect accounts of diagnostic events in respondents’ own words, drawing on established methods for narrative elicitation.32,33 The survey also collected data on the perceived sequelae of the diagnostic mistakes and problems, including physical impacts,34 financial impacts,35 and emotional impacts,36 all known outcomes for medical errors, but less studied for diagnostic-specific P&Ms.
Our technical expert panel, which included patient representatives and researchers with expertise in patient-reported diagnostic events, reviewed and endorsed the terminology used. We then pilot-tested the survey on a sample of 671 respondents, using the open-ended descriptions of diagnostic experiences to verify that events characterized as “problems” and “mistakes” both represented episodes meriting attention for diagnostic improvement (see Supplement 1 for additional detail, Supplemental Digital Content 1, http://links.lww.com/MLR/D141).
Source of Data and Survey Sample
We fielded the population-based survey drawing on the NORC AmeriSpeak Panel’s nationally representative sample between March and May 2023. This probability-based panel provides sample coverage of ~97% of the US household population. AmeriSpeak panelists were randomly selected for participation in this study. Those who were invited were told only that they would be surveyed about their health care experiences. Eligible participants were patients aged 18 or older and care partners (family or other individuals with a close relationship to a patient in their “household”).
The survey was administered on the web. NORC applied cleaning rules to the survey data for quality control: removing speeders (ie, those who completed the survey in less than one-third of the median duration) and removing respondents with high refusal rates (ie, those who skipped or refused > 50% of the eligible questions). The resulting survey response rate was 26.5%, with 95.4% completing the entire set of questions. The following demographic data were collected about each participant: age, gender (including nonbinary), race/ethnicity, and education.
Survey Questions
Participants were first asked whether they or someone close to them experienced any diagnostic problem or mistake in the past 4 years. Extended lookback periods are common for surveys of patient experiences involving safety events, ranging from 1 to 10 years.14,16,18,19 Participants who responded “yes” were then asked questions regarding those experiences, including how they came to know a diagnostic P&M had occurred, whether it was acknowledged by their doctor where the P&M occurred, and the physical, financial, and emotional impact of the diagnostic P&M (see Supplement 2, Supplemental Digital Content 1, http://links.lww.com/MLR/D141 for survey invitation and Supplement 3 for survey questions, Supplemental Digital Content 1, http://links.lww.com/MLR/D141).
Participants who reported experiencing multiple P&Ms were guided to select a single P&M for the survey responses. If they experienced 1 or more P&Ms involving themselves, as well as any involving a member of their household, they were asked to respond about their own experiences as a patient. Participants who reported experiencing multiple diagnostic P&Ms in the past 4 years were then instructed to reflect only on their most memorable P&M when responding to questions. The open-ended questions asked participants to describe the nature and impact of the diagnostic P&M. For the physical health item, which included death, respondents were asked: “when the diagnostic mistakes or problems were happening, how was (your/their) physical health affected overall? Did their physical health (1) stay the same; (2) get somewhat worse; (3) get much worse; or (4) did they die?” Only care partner respondents were given option 4 (“did they die?”).
Participants were asked to classify the diagnostic P&M into 1 of 3 categories: “clearly a mistake or mistakes,” “a problem or problems, not involving any mistakes,” or “both problems and mistakes.” We henceforth refer to these 3 categories as: “mistake(s),” “problem(s),” or “problems plus mistakes.”
The survey went through multiple phases of pilot testing and triangulation with qualitative data from a subset of respondents who completed more extensive personal interviews. When responses to the narrative questions did not fully match the scope of the responses from the interviews, we revised the wording of the narrative questions to encourage responses that filled these “gaps.”
Cleaning the Sample Using Narrative Content
Assessments of accompanying participant narratives, in which the participant described the diagnostic P&M, were conducted by at least 2 expert qualitative researchers, including 1 physician. Narratives in which the respondent was describing concerns about their care or treatment that did not explicitly involve the diagnostic process were excluded from the sample (6.2%). In these excluded cases, the participants typically described a treatment mishap unrelated to diagnosis, such as a surgical mistake or medication error.
Statistical Analyses
We calculated summary statistics for respondents who experienced a diagnostic P&M compared with those who did not. We compared characteristics of reported diagnostic P&Ms between patients and care partners, and over the 4-year lookback period. Respondents were also asked if their diagnosis had been affected by the pandemic. We conducted a sensitivity analysis to assess whether excluding respondents who reported that COVID had affected their diagnostic experiences altered the comparisons between problems and mistakes.
We also conducted χ2 analyses to compare participant responses to questions describing the impact and clinician involvement in the event across those who reported problems, those who reported mistakes, and those who reported problems plus mistakes. The questions about the emotional and financial impact of the diagnostic P&M asked the respondent to report on behalf of themselves, regardless of whether they were the patient or the care partner, whereas the questions on physical impact asked the care partner to respond on behalf of the patient.
Ethical Review
The study was determined to be exempt by the institutional review boards at REDACTED.
RESULTS
Sample Demographic Characteristics
Of the 3684 survey respondents, about one-third (32.9%) were over 60, 22.5% were between the ages of 45–59, 28.1% were between the ages of 30–44, and 16.5% of the sample were 18–29 (Table 1). About half of the overall respondents were female [1865/3684 (50.6%)]. The majority [2307/3684 (62.6%)] identified as White, 12.8% identified as Black, and 17.9% as Hispanic. Fewer than half had a college degree [1268/3684 (34.4%)].
TABLE 1.
Comparison of Key Respondent Characteristics of Those Who Reported a Diagnostic Problem or Mistake and Those Who Did Not
| Experience with diagnostic problems and mistakes in the past 4 y |
||||
|---|---|---|---|---|
| Respondent characteristics | Overall, n (%) | No problem or mistake, n (%) | Problem, mistake, or problem plus mistake, n (%) | Pearson χ2, P |
|
| ||||
| Overall | 3684 | 2468 (67.0) | 1216 (33.0) | — |
| Age | ||||
| 18–29 | 607 (16.5) | 332 (13.5) | 275 (22.6) | 107.2, < 0.001 |
| 30–44 | 1036 (28.1) | 651 (26.4) | 385 (31.7) | — |
| 45–59 | 829 (22.5) | 549 (22.2) | 280 (23.0) | — |
| > 60 | 1212 (32.9) | 936 (37.9) | 276 (22.7) | — |
| Gender | ||||
| Female | 1865 (50.6) | 1231 (49.9) | 582 (47.9) | 1.7, 0.20 |
| Male | 1819 (49.4) | 1237 (50.1) | 634 (52.1) | — |
| Nonbinary | 21 (0.58) | 9 (0.4) | 12 (1.0) | — |
| Race/ethnicity | ||||
| White non-Hispanic | 2307 (62.6) | 1610 (65.2) | 697 (57.3) | 21.8, 0.001 |
| Black non-Hispanic | 473 (12.8) | 300 (12.2) | 173 (14.2) | 3.1, 0.08 |
| Hispanic | 661 (17.9) | 409 (16.6) | 252 (20.7) | 9.5, 0.002 |
| Other | 150 (4.1) | 98 (4.0) | 52 (4.4) | 0.19, 0.66 |
| Education | ||||
| No high school degree | 271 (7.4) | 167 (6.8) | 104 (7.4) | 8.6, 0.073 |
| High school degree | 731 (19.8) | 497 (20.1) | 234 (19.2) | — |
| Some college | 1414 (38.4) | 925 (37.5) | 489 (40.2) | — |
| College degree | 731 (19.8) | 507 (20.5) | 224 (18.4) | — |
| Postgrad study/degree | 537 (14.6) | 372 (15.1) | 165 (13.6) | — |
About a third [1216/3684 (33.0%)] of screened households reported a diagnostic P&M in the past 4 years involving either themselves [697/3684 (18.9%)] or someone close to them [519/3684 (14.1%)]. Older respondents were less likely to report a diagnostic P&M than younger respondents (P < 0.001). There was no significant difference by gender in those who reported a diagnostic P&M and those who did not. Respondents who identified as Hispanic were significantly more likely to have experienced a diagnostic P&M (P < 0.05); those who identified as White non-Hispanic were less likely (P < 0.05).
Overall Prevalence of Diagnostic Problems and Mistakes
Of the 1216 respondents reporting a diagnostic P&M, the majority [697/1216 (57.3%)] reported that the P&M happened to them. The remaining respondents reported the P&M happened to their spouse [133/1216 (10.9%)], a family member who lives in their home [156/1216 (12.8%)], a family member who lives outside their home [150/1216 (12.3%)], a person living in their home who is not related to them [31/1216 (2.5%)], or a person living outside their home who is not related to them [46/1216 (3.8%)].
Characteristics of Diagnostic Problems and Mistakes Reported Overall
Across all screened households, over a tenth [448/3684 (12.2%)] reported that they or someone close to them only experienced a mistake, a tenth reported they only experienced a problem [382/3684 (10.4%)], and a tenth reported they experienced problems plus mistakes [371/3684 (10.1%)]. The comparable proportions for those reporting a P&M that happened to themselves were 5.8% (213/3684), 7.4% (271/3684), and 5.5% (203/3684).
Over a third of respondents [432/1216 (35.5%)] said the diagnostic P&M was acknowledged by the health care team working where the P&M occurred. About one-fourth of respondents first came to know of the diagnostic P&M because a health care professional where the P&M occurred told them [285/1216 (23.4%)]; 16.6% (200/1216) were told by a health care professional unaffiliated with the location where the P&M occurred. Most diagnostic P&Ms were reported to occur, at least in part, at either doctors’ offices (66.6%) or emergency departments (33.5%).
Over half of respondents reported the physical health of the patient got worse during the diagnostic P&M [714/1216 (58.7%)] and that their household finances were impacted [830/1216 (68.3%)]. Most respondents were emotionally impacted shortly after their discovery of the diagnostic P&M [1072/1216 (88.2%)], with about a third [414/1216 (34.3%)] reporting feelings of depression.
Patient Versus Care Partner Variations and Bias Related to Recall and COVID
Reports of worsening physical health among patients following the diagnostic P&M did not significantly differ between patients and care partners (Table 2). Supplement 4, Supplemental Digital Content 1, http://links.lww.com/MLR/D141 depicts comparisons of number of P&Ms experienced, short-term severe physical impact, whether the respondent was able to recall the severity of physical consequences, and length of accompanying narratives based on year of the P&M. None of the comparisons yielded any statistically significant cross-year differences, nor did they demonstrate any sort of persistent trend when comparing the most recent year to those that began more than 4 years in the past. Of the 14.7% (179/1216) who reported that the P&M was affected by COVID-19, there was no significant difference in the prevalence of problems, mistakes, or problems plus mistakes (P = 0.85; Supplement 5, Supplemental Digital Content 1, http://links.lww.com/MLR/D141) compared with unaffected respondents.
TABLE 2.
Characteristics and Impact by Patient Versus Care Partner Who Reported Experiencing a Diagnostic Problem and/or Mistake
| Respondent characteristics | Patients, n (%) | Care partners, n (%) | Pearson χ2, P |
|---|---|---|---|
|
| |||
| Age | |||
| 18–29 | 138 (19.8) | 137 (26.4) | 15.0, 0.002 |
| 30–44 | 218 (31.3) | 167 (32.2) | — |
| 45–59 | 158 (22.7) | 122 (23.5) | — |
| > 60 | 183 (26.3) | 93 (17.9) | — |
| Gender | |||
| Female | 1231 (49.9) | 582 (47.9) | 1.7, 0.20 |
| Male | 1237 (50.1) | 634 (52.1) | — |
| Nonbinary | 9 (0.4) | 12 (1.0) | — |
| Race/ethnicity | |||
| White non-Hispanic | 414 (59.4) | 283 (54.5) | 2.88, 0.09 |
| Black non-Hispanic | 96 (13.8) | 77 (14.8) | 0.28, 0.60 |
| Hispanic | 130 (18.6) | 122 (23.5) | 4.27, 0.04 |
| Education level | |||
| No high school degree | 56 (8.0) | 48 (9.2) | 8.6, 0.073 |
| High school degree | 119 (17.1) | 115 (22.2) | — |
| Some college | 297 (42.6) | 192 (37.0) | — |
| College degree | 132 (18.9) | 92 (17.7) | — |
| Postgrad study/degree | 93 (13.3) | 172 (13.9) | — |
| Perceived emotional and physical impact | |||
| Physical health was much worse | 120 (17.2) | 91 (17.5) | 0.02, 0.88 |
| Patient died | NA | 54 (10.4) | NA |
| Physical health was extremely/strongly impacted | 147 (21.1) | 111 (21.4) | 0.01, 0.90 |
| Household finances were affected by increased medical expenses | 460 (66.0) | 370 (71.3) | 3.84, 0.05 |
| Emotions experienced: sadness | 258 (37.0) | 205 (39.5) | 0.71, 0.40 |
| Emotions experienced: anger | 324 (46.5) | 265 (51.0) | 2.36, 0.12 |
| Emotions experienced: anxiety | 407 (58.5) | 244 (47.0) | 16.04, < 0.001 |
| Emotions experienced: depression | 260 (37.3) | 154 (29.7) | 7.94, 0.005 |
| Emotions experienced: helpless | 320 (45.9) | 237 (45.7) | 0.02, 0.90 |
| Reported at least 1 emotional response (sadness, anger, anxiety, depression, and helplessness) | 566 (81.2) | 427 (82.3) | 0.23, 0.63 |
Differences Between Mistakes and Problems Affecting Diagnostic Outcomes
There were no significant differences in the age, race and ethnicity, and education level of the respondents who reported experiencing a problem, a mistake, or problems plus mistakes, but women were more likely to report problems plus mistakes (59.0% vs. 50.0%, P = 0.003; Table 3). When care partner respondents reported problems plus mistakes, nearly twice as many reported the person died as a result [26/168 (15.5%)] compared with those who experienced only a problem [11/169 (6.5%)] or only a mistake [17/177 (9.6%)] (P = 0.016; Table 4). When patients survived, respondents reporting problems plus mistakes were more likely to indicate that physical health had become much worse [84/371 (39.8%)] compared with those reporting only problems [53/382 (25.1%)] or only mistakes [72/448 (34.1%)] (P = 0.002).
TABLE 3.
Comparison of Sociodemographic Characteristics of Respondents Reporting Problems, Mistakes, and Problems Plus Mistakes*
| Respondent characteristics | Only problem(s), n (%) | Only mistake(s), n (%) | Problem(s) plus mistake(s), n (%) | Pearson χ2, P |
|---|---|---|---|---|
|
| ||||
| Overall | 382 (31.4) | 448 (36.8) | 371 (30.5) | — |
| Age | ||||
| 18–29 | 85 (22.6) | 111 (24.8) | 74 (19.2) | 4.33, 0.63 |
| 30–44 | 117 (30.6) | 144 (32.1) | 122 (32.9) | — |
| 45–59 | 94 (24.6) | 98 (21.9) | 83 (22.4) | — |
| > 60 | 86 (22.5) | 95 (21.2) | 92 (24.8) | — |
| Gender† | ||||
| Female | 199 (52.1) | 211 (47.1) | 219 (59.0) | 11.60, 0.003 |
| Male | 183 (47.9) | 237 (52.9) | 152 (41.0) | — |
| Nonbinary | 5 (1.3) | 4 (0.9) | 3 (0.8) | — |
| Race/ethnicity | ||||
| White non-Hispanic | 219 (57.3) | 245 (54.7) | 228 (61.5) | 3.82, 0.15 |
| Black non-Hispanic | 60 (15.7) | 68 (15.2) | 42 (11.3) | 3.60, 0.17 |
| Hispanic | 74 (19.3) | 102 (22.8) | 69 (18.6) | 2.54, 0.281 |
| Other | 21 (4.7) | 16 (4.2) | 15 (4.0) | 0.23, 0.89 |
| Education | ||||
| No high school degree | 29 (7.6) | 44 (9.8) | 27 (7.3) | 6.27, 0.617 |
| High school degree | 77 (20.2) | 81 (18.1) | 70 (18.9) | — |
| Some college | 141 (36.9) | 187 (41.7) | 157 (42.3) | — |
| College degree | 75 (19.6) | 78 (17.4) | 70 (18.9) | — |
| Postgrad study/degree | 60 (15.7) | 58 (13.0) | 47 (12.7) | — |
| Who experienced the diagnostic problem and/or mistake that happened to | ||||
| The respondent | 213 (55.8) | 271 (60.5) | 203 (54.7) | 3.24, 0.198 |
| Care partner | 169 (44.2) | 177 (39.5) | 168 (45.3) | — |
| Relationship of the care partner to the patient | ||||
| Spouse | 56 (33.1) | 43 (24.3) | 32 (19.2) | 22.86, 0.004 |
| Family member who lives in the home | 52 (30.8) | 64 (36.2) | 40 (24.0) | — |
| Family member who lives outside the home | 35 (20.7) | 48 (27.1) | 66 (40.0) | — |
| A person who lives in a home but is not related | 10 (5.9) | 9 (5.1) | 12 (7.2) | — |
| A person who lives outside your home and is not related | 16 (9.5) | 13 (7.3) | 17 (10.2) | — |
This table presents the breakdown of the percent of people who reported “yes” to experiencing a diagnostic problem or mistake (n = 1216) and further classified the P&M as a problem, mistake, or both (n = 1201), so the denominator is 1201. Fifteen participants who did not provide a response to the item classifying their P&M were also excluded from this analysis.
Gender was collected through self-report.
TABLE 4.
Participants’ Experiences and Perceptions of Reported Diagnostic P&Ms Across Problems and Mistakes, and the Health Care Team’s Response
| Patient Experiences and Perceptions | Only problem(s) n = 382, n (%) | Only mistake(s) n = 448, n (%) | Pearson χ2 (only problems and only mistakes), P | Problems plus mistake(s) n = 371, n (%) | Pearson χ 2 (comparing all 3 categories), P |
|---|---|---|---|---|---|
|
| |||||
| Physical health was much worse* | 53 (25.1) | 72 (34.1) | 0.8, 0.38 | 84 (39.8) | 12.8, <0.001 |
| Patient died† | 11 (2.9) | 17 (3.8) | 0.5, 0.47 | 26 (7.1) | 8.29, 0.016 |
| Physical health was extremely/strongly impacted | 59 (15.4) | 80 (17.9) | 0.9, 0.35 | 119 (32.1) | 36.4, <0.001 |
| Household finances were affected by increased medical expenses | 244 (63.9) | 299 (66.7) | 0.8, 0.39 | 285 (76.8) | 18.5, <0.001 |
| Emotions experienced: sadness | 127 (33.5) | 149 (33.3) | 0, 0.96 | 185 (50.0) | 29.7, <0.001 |
| Emotions experienced: anger | 128 (33.8) | 224 (50.1) | 22.4, <0.001 | 235 (63.5) | 66.6, <0.001 |
| Emotions experienced: anxiety | 192 (51.4) | 222 (49.7) | 0.26, 0.61 | 233 (63.0) | 16.3, <0.001 |
| Emotions experienced: depression | 111 (29.3) | 149 (33.3) | 1.6, 0.21 | 153 (41.4) | 12.5, 0.002 |
| Emotions experienced: helpless | 176 (46.4) | 165 (36.9) | 7.7, <0.01 | 216 (58.4) | 37.5, <0.001 |
| Reported at least 1 emotional response (sadness, anger, anxiety, depression, or helplessness) | 293 (76.7) | 367 (81.9) | 2.29, 0.13 | 330 (88.9) | 19.6, <0.001 |
| Health care team’s response | |||||
| A health care professional at the diagnostic site informed the respondent about P&M‡ | 97 (25.4) | 101 (22.5) | 0.9, 0.33 | 86 (23.2) | 0.99, 0.61 |
| Health care team acknowledged diagnostic P&M§ | 131 (34.3) | 180 (40.2) | 3.0, 0.082 | 118 (31.8) | 6.9, 0.032 |
| Received an apology from the health care team‖ | 86 (44.8) | 123 (35.2) | 69.7, <0.001 | 63 (22.8) | 25.6, <0.001 |
This analysis was limited to the subsample of patients who had survived until the time the survey was fielded. This resulted in the following sample sizes: “only problems”—n = 371, “only mistakes”—n = 431, “problems plus mistakes”—n = 287. Fifteen participants who did not provide a response to the item classifying their P&M were also excluded from this analysis.
Only care partner respondents were given the option to report death. While the percentages shared are based on the overall denominator for clarity, when we limit to care partner respondents, the percentages are: problems [11/169 (6.5%)], mistakes [17/177 (9.6%)], and problems plus mistakes [26/168 (15.5%)].
Respondents reported how they first discerned that there had been a diagnostic problem or mistake.
Respondents were asked this item regardless of their response to whether the health care professional at the diagnostic site informed the respondent about the P&M.
We excluded participants who reported that they did not think that an apology was warranted, given the nature of their diagnostic P&M. This resulted in sample sizes of: “only problems”—n = 192, “only mistakes”—n = 349, “problems plus mistakes”—n = 276.
A similar pattern emerged for other reported outcomes: problems plus mistakes were associated with more severe consequences than either mistakes alone or problems alone. The majority of participants [330/371 (88.9%)] reported at least 1 emotion of sadness, anger, anxiety, depression, or helplessness as a result of the problems plus mistakes (P < 0.001). While household finances were affected by medical expenses for almost all P&Ms, those who reported problems plus mistakes were most likely to report these financial consequences (problems plus mistakes: 61.4%, problems: 46.0%, mistakes: 50.3%, P < 0.001).
Equally striking is the lack of a consistent difference in harms between singular mishaps that respondents classified as “problems” versus those that they reported as harms. There were no statistically significant differences in terms of health effects or financial burdens. The emotional impacts differed in terms of the specific affect they triggered—mistakes more frequently evoked anger, whereas problems were more often associated with a sense of helplessness—but the overall prevalence of negative emotional responses was equivalent for problems and mistakes.
Differences in the Health Care Team’s Response to the Reported Diagnostic P&M
There were no significant differences across types of P&Ms in whether a respondent came to know a P&M had occurred because a health care team member where the P&M occurred told them (P = 0.61; Table 4). There were differences in terms of whether the P&M was acknowledged by the health care team where the P&M occurred [problems: 131/382 (34.3%), mistakes: 180/448 (40.2%), both: 118/371 (31.8%), P = 0.032]. However, these were statistically significant only in terms of the reduced acknowledgment when problems were combined with mistakes: acknowledgment was equally common for singular problems compared with singular mistakes.
Excluding those who reported that an apology was not necessary (n = 817), respondents who reported a diagnostic mistake or a problem plus mistake were less likely to report that they received an apology from the health care team where the P&M occurred compared with those who reported only a problem [problems: 86/192 (44.8%), mistakes: 123/349 (35.2%), problems plus mistakes: 63/276 (22.8%), P < 0.001]. Taken together, the 3 metrics suggest that providers were seeing events reported by our respondents as “problems” to be as or more consequential than those reported as “mistakes.”
DISCUSSION
This study offers new insights into how patients and care partners perceive and report their diagnostic experiences when given the more inclusive framing of problems and mistakes (P&Ms), rather than being asked to fit their experiences into a single category of “diagnostic error.”37–39 Our findings further suggest that (1) the prevalence of diagnostic P&Ms reported by the public is considerably higher when defined as involving problems and/or mistakes than when restricted to mistakes alone, and (2) the additional cases included in the “problems” or “combination” categories are as impactful on patient well-being as those classified exclusively as “mistakes.” The majority of prior studies measuring diagnostic breakdowns have relied primarily on sources that do not include direct input from patients, such as electronic health records (EHRs) and other administrative data.39–41 Our findings provide yet another strong justification for attention to a more inclusive set of diagnostic experiences to improve diagnostic quality.
Perhaps the most striking set of findings from our study comes from the group of respondents who classified their diagnostic experiences as involving a combination of problems plus mistakes. We hypothesize that P&Ms classified as involving problems plus mistakes were the most likely to involve a complex diagnostic trajectory. We found that when respondents identified the diagnostic P&M as involving problems plus mistakes, they were nearly twice as likely to report death as a result of the diagnostic P&M. P&Ms classified as involving problems plus mistakes, however, were the least likely to be acknowledged by the health care team where the P&M occurred. This disjunction between harms and acknowledgment may reflect heightened concerns about litigation. Alternatively, because P&Ms categorized as problems plus mistakes were more likely than other P&Ms to emerge across multiple settings (59% vs. 48%), the greater number of providers involved may diffuse perceived responsibility.
The results of this national survey demonstrate the analytical richness of a patient-oriented and inclusive definition of diagnostic P&Ms for use with both patients and care partners. Since we found that this broader definition—asking for descriptions of experiences with “problems and mistakes”—revealed a high prevalence of patient-reported diagnostic P&Ms, our study also suggests that including the term “diagnostic problems” seemed to encourage sharing from participants who were unsure of the diagnostic P&M’s cause, as they were significantly less likely to report that they felt they were owed an apology, though they still experienced a significant level of physical, emotional, and financial harm. The reported impact of diagnostic events categorized as problems appears far too substantial to ignore and is highly valuable to the primary care workforce, who advocate vociferously for patients and their care partners.
Only the patient and their care partner(s) can provide insight into the entirety of the diagnostic process, from when their symptoms started, to when and if they received a diagnosis, to how their subsequent health trajectory unfolded.42,43 Using inclusive, patient-centered language is essential to help individuals reflect on their diagnostic experience through a broad lens, enabling the identification of critical system-level issues that may otherwise be overlooked when the framing is too narrow or when data collection excludes quantifiable patient input.44 Because diagnosis is often a longitudinal process, what is ultimately perceived as an “error” may stem from multiple overlapping factors over time.20,44,45
As attention is increasingly paid to the value of diagnostic excellence,44,46,47 a concomitant recognition of patients and their care partners as experts who can offer insights into system-level issues and root causes of diagnostic issues is warranted. Patients and their care partners can also shed light on whether factors largely outside the clinical team’s control—including cost, insurance coverage, and resources to attend health care visits—impacted their diagnostic process and its outcomes. For instance, nearly two-thirds of respondents reported that their household finances were affected by increased medical expenses related to diagnostic P&Ms, a substantial impact. Future work that examines the actionability of accompanying patient narratives to diagnostic P&Ms has the potential to guide quality improvement at the clinic level and shape policy responses. To make this feasible, feedback arrangements must assist clinicians and practice administrators in interpreting and responding to patient-reported data and narratives.22 Recent studies document the feasibility and impact of such feedback arrangements in clinical settings,9,48,49 but they have not yet been applied to diagnostic improvement.
Potential Limitations
The participation rate in the survey was 26.5%. This may affect the generalizability of results, though the choice to participate was made before panelists learned that the survey asked about diagnostic experiences, eliminating bias in these terms. Respondents mostly reflected the general US population with some modest differences: there were more adults aged 60 or older (34.8% vs. 28.4%),50 fewer adults with an income of $100,000 or above (24.7% vs. 33.6%),51 more people with a college education or greater (34.1% vs. 29.7%),52 and a similar distribution across race and ethnicity.
The 4-year time span covered by our survey encompassed the COVID-19 pandemic; as a result, responses may not reflect typical care.53–55 A little less than a third of respondents reported that their diagnostic problem was affected by the pandemic. However, excluding these respondents from our analyses did not alter the comparisons between problems and mistakes.
We included care partners as reporters and did not require a familial relationship. This inclusion may create a concern that these respondents would be less knowledgeable than patient respondents about the details of the diagnostic process or its subsequent sequelae. However, two-thirds of the care partners reported going with the patients to their medical appointments where the diagnostic P&M occurred. We also found that the reported prevalence and key characteristics of diagnostic P&Ms did not differ between patients and care partners; thus, excluding care partners from the sample would not change key findings.
Finally, we used an extended lookback period of 4 years. Other efforts to elicit patient-reported diagnostic P&Ms used similar time frames.18 However, health experiences evolve over extended time periods and may influence respondents’ assessments in ways not yet fully assessed.
Future Research Directions
Future research can further enhance our understanding by more fully examining the downstream consequences and system-level determinants of patient- and care partner-reported diagnostic P&Ms. In particular, rigorous investigation of the financial impact of diagnostic P&Ms is needed to quantify both direct medical costs (eg, repeat consultations, diagnostic testing, and delayed treatment) and indirect burdens (eg, lost wages, caregiving time, transportation, and insurance instability) borne by patients and families. This study did not collect information on insurance coverage or the frequency of health care visits, both of which are likely to mediate these financial impacts. By so doing, coverage also influences patients’ perceived risk of experiencing a harmful diagnostic problem or mistake. Future work should assess how these factors shape patients’ diagnostic experiences and associated financial strain. Finally, future research should more explicitly evaluate the role of care partners within the diagnostic process, including how their observations, advocacy behaviors, and communication with clinicians influence diagnosis timeliness, accuracy, and escalation of concern.
CONCLUSION
Patients and care partners report diagnostic P&M experiences, whether labeled as problems or mistakes, with substantial emotional, physical, and financial impact. These findings underscore the importance of designing survey instruments with inclusive language to better capture patient feedback on diagnosis. They also point to the need for institutional mechanisms that translate this feedback into actionable diagnostic quality improvement efforts. Terms like “problems” and “mistakes” may be more effective for routine use in feedback systems, particularly when aiming to engage diverse populations across a range of clinical settings. Capturing these nuanced perspectives can guide improvements in diagnostic quality within learning health systems.
Supplementary Material
Supplemental Digital Content is available for this article. Direct URL citations are provided in the HTML and PDF versions of this article on the journal’s website, www.lww-medicalcare.com.
Acknowledgments
The authors declare that financial support was received for the research, authorship, and/or publication of this article. This study was supported by AHRQ R18HS029350 (PIs McDonald & Newman-Toker), AHRQ R01HS029789 (PI Schlesinger (Yale); Hopkins Subcontract PI McDonald; U of Wisconsin Subcontract PI Grob), Gordon and Betty Moore Foundation Grant number 10781 (PI Schlesinger (Yale); Hopkins Subcontract PI McDonald; U of Wisconsin Subcontract PI Grob).
Footnotes
The authors declare no conflict of interest.
REFERENCES
- 1.Berwick DM, Nolan TW, Whittington J. The triple aim: care, health, and cost. Health Aff (Millwood). 2008;27:759–769. [DOI] [PubMed] [Google Scholar]
- 2.Navarro V Assessment of the World Health Report 2000. Lancet. 2000;356:1598–1601. [DOI] [PubMed] [Google Scholar]
- 3.Minvielle E, Fierobe A, Fourcade A, et al. The use of patient-reported outcome and experience measures for health policy purposes: a scoping review in oncology. Health Policy. 2023;129:104702. [DOI] [PubMed] [Google Scholar]
- 4.Snowdon DA, Srikanth V, Beare R, et al. A landscape assessment of the use of patient reported outcome measures in research, quality improvement and clinical care across a healthcare organisation. BMC Health Serv Res. 2023;23:94. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Knowles SE, Allen D, Donnelly A, et al. Participatory codesign of patient involvement in a learning health system: how can data-driven care be patient-driven care? Health Expect. 2022;25:103–115. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Grob R, Gleason K, McLean P, et al. Patients’ roles in governance of learning: results from a qualitative study of 16 learning healthcare systems. Learn Health Syst. 2022;6:e10269. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Harrison R, Walton M, Manias E, et al. The missing evidence: a systematic review of patients’ experiences of adverse events in health care. Int J Qual Health Care. 2015;27:424–442. [DOI] [PubMed] [Google Scholar]
- 8.Birkeland S Health care complaints and adverse events as a means of user involvement for quality and safety improvement. Milbank Q. 2019;97:346–349. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Schlesinger M, Grob R, Shaller D. Using patient-reported information to improve clinical practice. Health Serv Res. 2015;50 (suppl 2):2116–2154. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Panagioti M, Khan K, Keers RN, et al. Prevalence, severity, and nature of preventable patient harm across medical care settings: systematic review and meta-analysis. Brit Med J. 2019;366:l4185. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Zhu J, Stuver SO, Epstein AM, et al. Can we rely on patients’ reports of adverse events? Med Care. 2011;49:948–955. [DOI] [PubMed] [Google Scholar]
- 12.De Brun A, Heavey E, Waring J, et al. PReSaFe: a model of barriers and facilitators to patients providing feedback on experiences of safety. Health Expect. 2017;20:771–778. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Mahmoud HA, Thavorn K, Mulpuru S, et al. Barriers and facilitators to improving patient safety learning systems: a systematic review of qualitative studies and meta-synthesis. BMJ Open Qual. 2023;12:e002134. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Ricci-Cabello I, Avery AJ, Reeves D, et al. Measuring patient safety in primary care: the development and validation of the “Patient Reported Experiences and Outcomes of Safety in Primary Care” (PREOS-PC). Ann Fam Med. 2016;14:253–261. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Rainey H, Ehrich K, Mackintosh N, et al. The role of patients and their relatives in “speaking up” about their own safety—a qualitative study of acute illness. Health Expect. 2015;18:392–405. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Bell SK, Bourgeois F, Dong J, et al. Patient identification of diagnostic safety blindspots and participation in “good catches” through shared visit notes. Milbank Q. 2022;100:1121–1165. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Williams-Bader J, McDonald KM, Drye EE. Advancing measurement of diagnostic excellence for better healthcare. Jt Comm J Qual Patient Saf. 2025;51:386–387. [DOI] [PubMed] [Google Scholar]
- 18.Institute for Healthcare Improvement; NORC at the University of Chicago. Americans’ experiences with medical errors and views on patient safety. 2017.
- 19.Kistler CE, Walter LC, Mitchell CM, et al. Patient perceptions of mistakes in ambulatory care. Arch Intern Med. 2010;170:1480–1487. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Yang D, Fineberg HV, Cosby K. Diagnostic excellence. JAMA. 2021;326:1905–1906. [DOI] [PubMed] [Google Scholar]
- 21.Dukhanin V, Gamper MJ, Gleason KT, et al. Patient-reported outcome and experience domains for diagnostic excellence: a scoping review to inform future measure development. Qual Life Res. 2024;33:2883–2897. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.McDonald KM, Gleason KT, Jojodia A, et al. Achieving diagnostic excellence: roadmaps to develop and use patient-reported measures with an equity lens. Int J Health Policy Manag. 2024;13:8048. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Rosenthal M, Schlesinger M. Not afraid to blame: the neglected role of blame attribution in medical consumerism and some implications for health policy. Milbank Q. 2002;80:41–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Schlesinger M, Mitchell S, Elbel B. Voices unheard: barriers to expressing dissatisfaction to health plans. Milbank Q. 2002;80:709–755. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Sattar R, Johnson J, Lawton R. The views and experiences of patients and health-care professionals on the disclosure of adverse events: a systematic review and qualitative meta-ethnographic synthesis. Health Expect. 2020;23:571–583. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Giardina TD, Haskell H, Menon S, et al. Learning from patients’ experiences related to diagnostic errors is essential for progress in patient safety. Health Aff (Millwood). 2018;37:1821–1827. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Schlesinger M The canary in the gemeinshaft: the public voice of patients as a means of enhancing health system performance, In: Hoffman B, Tomes N, Grob R, et al. eds. Patients as Policy Actors. Rutgers University Press; 2011:148–176. [Google Scholar]
- 28.Apodaca C, Casanova-Perez R, Bascom E, et al. Maybe they had a bad day: how LGBTQ and BIPOC patients react to bias in healthcare and struggle to speak out. J Am Med Inform Assoc. 2022;29:2075–2082. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Bell SK, Bourgeois F, DesRoches CM, et al. Filling a gap in safety metrics: development of a patient-centred framework to identify and categorise patient-reported breakdowns related to the diagnostic process in ambulatory care. BMJ Qual Saf. 2022;31:526–540. [DOI] [PubMed] [Google Scholar]
- 30.Kuzel AJ, Woolf SH, Gilchrist VJ, et al. Patient reports of preventable problems and harms in primary health care. Ann Fam Med. 2004;2:333–340. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Ward JK, Armitage G. Can patients report patient safety incidents in a hospital setting? A systematic review. BMJ Qual Saf. 2012;21:685–699. [DOI] [PubMed] [Google Scholar]
- 32.Grob R, Schlesinger M, Parker AM, et al. Breaking narrative ground: innovative methods for rigorously eliciting and assessing patient narratives. Health Serv Res. 2016;51(suppl 2):1248–1272. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Martino SC, Reynolds KA, Grob R, et al. Evaluation of a protocol for eliciting narrative accounts of pediatric inpatient experiences of care. Health Serv Res. 2023;58:271–281. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Newman-Toker DE, Nassery N, Schaffer AC, et al. Burden of serious harms from diagnostic error in the USA. BMJ Qual Saf. 2024;33:109–120. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Andel C, Davidow SL, Hollander M, et al. The economics of health care quality and medical errors. J Health Care Finance. 2012;39:39–50. [PubMed] [Google Scholar]
- 36.Prentice JC, Bell SK, Thomas EJ, et al. Association of open communication and the emotional and behavioural impact of medical error on patients and families: state-wide cross-sectional survey. BMJ Qual Saf. 2020;29:883–894. [DOI] [PubMed] [Google Scholar]
- 37.Singh H, Meyer AN, Thomas EJ. The frequency of diagnostic errors in outpatient care: estimations from three large observational studies involving US adult populations. BMJ Qual Saf. 2014;23:727–731. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Singh H, Giardina TD, Meyer AN, et al. Types and origins of diagnostic errors in primary care settings. JAMA Intern Med. 2013;173:418–425. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Newman-Toker DE, Schaffer AC, Yu-Moe CW, et al. Serious misdiagnosis-related harms in malpractice claims: the “big three”—vascular events, infections, and cancers. Diagnosis (Berl). 2019;6:227–240. [DOI] [PubMed] [Google Scholar]
- 40.Omron R, Kotwal S, Garibaldi BT, et al. The diagnostic performance feedback “calibration gap”: why clinical experience alone is not enough to prevent serious diagnostic errors. AEM Educ Train. 2018;2:339–342. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Newman-Toker DE, Moy E, Valente E, et al. Missed diagnosis of stroke in the emergency department: a cross-sectional analysis of a large population-based sample. Diagnosis. 2014;1:155–166. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Gleason KT, Peterson S, Dennison Himmelfarb CR, et al. Feasibility of patient-reported diagnostic errors following emergency department discharge: a pilot study. Diagnosis (Berl). 2021;8:187–192. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.McDonald KM, Bryce CL, Graber ML. The patient is in: patient involvement strategies for diagnostic error mitigation. BMJ Qual Saf. 2013;22(suppl 2):ii33–ii39. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Edlow JA, Pronovost PJ. Misdiagnosis in the emergency department: time for a system solution. JAMA. 2023;329:631. [DOI] [PubMed] [Google Scholar]
- 45.Yousef EA, Sutcliffe KM, McDonald KM, et al. Crossing academic boundaries for diagnostic safety: 10 complex challenges and potential solutions from clinical perspectives and high-reliability organizing principles. Hum Factors. 2022;64:6–20. [DOI] [PubMed] [Google Scholar]
- 46.AHRQ. Expands its Repertoire to Eliminate Diagnostic Errors press release. 2022.
- 47.Washington (DC): The National Academies of Sciences, Engineering, and Medicine; 2015. Improving Diagnosis in Health Care. Washington, DC: The National Academies Press. 10.17226/21794 [DOI] [PubMed] [Google Scholar]
- 48.Grob R, Lee YSH, Shaller D, et al. Nothing is more powerful than words:” how patient experience narratives enable improvement. Qual Manag Health Care. 2024;33:149–159. [DOI] [PubMed] [Google Scholar]
- 49.Shaller D, Nembhard I, Matta S, et al. Assessing an innovative method to promote learning from patient narratives: findings from a field experiment in ambulatory care. Health Serv Res. 2024;59:e14245. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.American Community Survey US Census Bureau. 2021.
- 51.Shrider E, Kollar M, Chen F, et al. Current Population Reports, P60–273, Income and Poverty in the United States. Office UGP. 2021. [Google Scholar]
- 52.National Health Interview Survey. 2022.
- 53.Carlicchi E, Di Sabato ME, Cincotta A, et al. How the COVID-19 pandemic affected attendance at a tertiary orthopedic center emergency department: a comparison between the first and second waves. Diagnostics (Basel). 2022;12:2855. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 54.Huang M, Khurana A, Mastorakos G, et al. Patient portal messaging for asynchronous virtual care during the COVID-19 pandemic: retrospective analysis. JMIR Hum Factors. 2022;9:e35187. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 55.Iglesias A, Ambrose A, Coronel-Mockler S, et al. Impact of transitioning to virtual delivery of a cardiovascular health improvement program for Latinos during the COVID-19 pandemic. BMC Public Health. 2022;22:1935. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.