Article, Radiology

Radiologic discrepancies in children with special healthcare needs in a pediatric emergency department

a b s t r a c t

Background: After-hours radiologic interpretation by nonradiology attendings or resident radiologists introduces the risk of discrepancies. Clinical outcomes following radiologic discrepancies among pediatric emergency de- partment (ED) patients are poorly described. In particular, children with special healthcare needs (CSHCN), have more opportunities for discrepancies and potential consequences than non- CSHCN. Our objective was to determine the rates and types of radiologic discrepancies, and to compare CSHCN to non-CSHCN.

Methods: From July 2014 to February 2015, all children who underwent a diagnostic Imaging study at a free- standing children’s ED were included. Data collected included Radiologic studies – type and location – and clinical details – chief complaint and CSHCN type. Differences between preliminary reads and final pediatric radiology at- tending reads were defined as discrepancies, and categorized by clinical significance. Descriptive statistics, z- tests, and chi-square were used.

Results: Over 8 months, 8310 visits (7462 unique patients) had radiologic studies (2620 CSHCN, 5690 non-

CSHCN). A total of 198 (2.4%) radiologic discrepancies [56 (28.3%) CSHCN, 142 (71.7%) non-CSHCN] were found. Chief complaints for CSCHN were more often within the cardiac, pulmonary and neurologic systems (p b 0.001 for each), whereas non-CSHCN presented with more trauma (p b 0.001). The rates of discrepancies (CSHCN 2.1%, non- CSHCN 2.5%, p = 0.3) and severity of clinical consequences (p = 0.6) were not significantly different between CSHCN and non-CSHCN.

Conclusion: Though the frequency and type of radiologic studies performed between CSHCN and non-CSHCN were different, we found no significant difference in the rate of radiologic discrepancies or the rate of clinically significant radiologic discrepancies.

(C) 2017

Introduction

radiologic imaging, such as radiographs (X-rays) and computed to- mography (CT) scans, is an important diagnostic tool for a range of pre- senting symptoms in the emergency department (ED). The current ED standard of care for off-hours radiology is for emergency medicine (EM) physicians and/or radiology trainees to perform their own prelim- inary interpretations of radiologic studies, which are subsequently reviewed by an attending radiologist the next day. An attending radiologist’s revision that overrides a preliminary read is considered a radiologic discrepancy, sometimes referred to as an ‘over-read.’ The lit- erature reflects discrepancy rates of 0.1% to 9.0% in adult populations [1-3] and a broader range of 0.1% to 28% in children [4-7]. Over-reads can result in Changes in management, including repeat imaging, special- ist follow up, and hospital admission [5].

* Corresponding author.

E-mail address: [email protected] (A. Festekjian).

Children with special healthcare needs (CSHCN) are medically com- plex patients with frequent ED encounters often involving radiologic imaging studies. Examples of disease processes that define patients as CSHCN include asthma, autism, diabetes mellitus, and cystic fibrosis. CSHCN represent a growing part of the pediatric population – an esti- mated 15-20% of US children age 0-17 – and have four times the num- ber of hospitalizations than other children [8]. With more interactions with the health care system for maintenance of their devices and chron- ic conditions, CHSCN have more opportunities to experience diagnostic errors and thus are likely to be at heightened risk for Negative outcomes due to those errors. The rate of radiologic discrepancies for CSHCN is currently unknown. CSHCN’s significant representation among total ED visits, especially in Tertiary care hospitals, can be high, and their clin- ical complexity justifies efforts to further investigate CSHCN specifically. The objective of this study was to report the rate of radiologic dis- crepancy in CSHCN, compared to that of the general pediatric ED popu- lation, and to determine differences in the types of radiologic diagnostic errors in CSHCN. Additionally, clinical consequences from radiologic

https://doi.org/10.1016/j.ajem.2017.12.041

0735-6757/(C) 2017

discrepancies were scored and compared between CSHCN and non- CSHCN. Given their complex medical conditions and more frequent ex- posure to Radiologic tests, it was hypothesized that the rate and impact of radiologic discrepancy for CSHCN is significantly greater than that of their peers.

Methods

Setting and population

From July 2014 to February 2015, all children who underwent a di- agnostic imaging study during their visit to the ED of an urban tertiary care free-standing children’s hospital were included in the study cohort. The pediatric emergency department (PED) at this hospital has over 80,000 unique visits annually, and up to 40% of patients seen in any part of the hospital are CSHCN. For this study, CSHCN were defined using the Maternal and Child Health Bureau definition, “those who have or are at increased risk for a chronic physical, developmental, be- havioral, or emotional condition and who also require health and relat- ed services of a type or amount beyond that required by children generally.” [9-10] The radiology department at this hospital performs N 26,000 radiologic studies every year for patients being evaluated in the PED. Pediatric radiologists are available to read diagnostic imaging studies during normal business hours, from approximately 8 am to 5 pm on weekdays. During Off hours, imaging studies are read by PED practitioners [board certified in Pediatrics and/or Pediatric Emergency Medicine (PEM)] and/or resident radiologists, and confirmed the fol- lowing morning by the attending radiologist. Resident radiologists re- view all studies until midnight and CT scans until the following morning. Radiologic discrepancies are discussed and documented daily with the PEM attending by the attending radiologist.

We defined radiologic discrepancies when the final interpretations of diagnostic imaging by the attending radiologist differed from the pre- liminary read by either the resident radiologist or PED practitioner. The clinical consequence of the discrepancy was then reviewed in the med- ical record and classified on the following 4-level scale, created specifi- cally for this study by the study team: no consequence (no change or already addressed in the PED), mild consequence (requiring a simple phone call such as new antibiotics or follow-up with a clinic), moderate consequence (requiring the patient to return to the ED for further evalu- ation or imaging), and high consequence (requiring immediate admis- sion, operation, or child abuse evaluation). Discrepancies were considered equivocal when a radiologic finding didn’t correlate clinical- ly (i.e. calcification in the right lower quadrant which could be appendi- citis in the right clinical setting). The study was approved by the Institutional Review Board.

We included any PED visit in which diagnostic imaging was per-

formed during the 8-month study period. Diagnostic images for this study were defined as studies that are commonly read by emergency physicians and radiology trainees: all plain X-rays were included (in- cluding studies with contrast), and computed tomography studies (CTs) of the brain or abdomen-pelvis, and ultrasounds of the abdomen, pylorus, and pelvis. Studies were excluded if they were performed on patients over the age of 18, or if they were ordered on patients who were not actively being treated in the PED. We excluded all other imag- ing not commonly interpreted by emergency physicians, such as mag- netic resonance imaging (MRI) and fluoroscopy. ED point-of-care ultrasounds were similarly excluded.

Data collection & variables

A trained research assistant (MF) reviewed all imaging studies per- formed on PED patients and, after reviewing the patient’s medical chart in the electronic medical record, categorized them into CSHCN or without SHCN using the definition cited above. Any conflicts on cat- egorization between CSHCN and children without SHCN were resolved

by investigator consensus (AF, DL, TC). Discrepancies were retrieved from a discrepancy log book which was kept as part of standard of care. These discrepancies were abstracted by the same trained research assistant and verified by a PEM investigator (AF, DL, TC). Clinical signif- icance of diagnostic imaging discrepancies was determined by a board- certified PEM physician (AF) and attending radiologist (HL), who was not the clinical radiologist on record. Discrepancies on patients admit- ted from the ED were not documented in the ED records and therefore excluded from this study.

Data were abstracted for factors at the patient and system levels. Patient-level variables included: age, sex, type of CSHCN, presence of medical equipment, tubes, or implants, the anatomic region imaged by the study and type of imaging study. Visit-level variables included: time of day, hospital census, and ED census as a proxy for ED crowding. Primary outcome variables included radiologic discrepancy and level of clinical consequence resulting from the discrepancy. Secondary out- come variables included types of chief complaints and radiologic studies performed.

Data analysis

The unit of analysis in this study was the PED visit encounter, not the patient, as both emergency physicians and radiologists treat each radio- logic study and visit independently. Descriptive statistics were used to summarize the data. Two-sample z-test and chi-square examined asso- ciations between imaging discrepancy and CSHCN status. Chi-square or Fisher Exact Test was employed to examine associations with CSHCN status and other visit-level variables and radiologic study. Continuous data as confounders were analyzed using point-biserial correlation. All analyses were two-tailed with an alpha of 0.05 and were conducted using IBM Statistical Package for the Social Sciences Version 23 (IBM SPSS, Armonk, NY).

Results

Descriptive statistics

Over the 8-month study period, 8310 discrete visits had radiologic studies, representing 7462 unique patients. Among the unique patients, 682 (9.1% of all patients) presented more than once during the 8-month period. The mean age at presentation was 8.0 +/- 5.8 years. CSHCN represented 2620 visits (31.5%) requiring radiographs. CSHCN patients tended to be older than non-CSHCN overall (9.1 vs. 7.5 years, p b 0.001), and a slightly higher proportion of CSHCN were male compared to non-CSHCN [56% (1460) vs 53% (3004), p = 0.01]. The clinical fea- tures and medical equipment characterizing CSHCN visits are listed in Table 1.

PED census ranged from 124 to 321 patients per day (mean 215.2

+/- 39.3). Only 226 of the 8310 visits (2.7%) were within 48 h of a prior PED visit.

Primary Question. From 8310 visits, there were a total of 198 (2.4%) documented radio-

logic discrepancies. Of these, 7 were classified as false positives and 134 as false negatives; the rest consisted of equivocal findings. Nearly half (100/203, 49.3%) required notification and a change in plan for the pa- tient (Table 2), but the clinical action required due to the discrepancy was not found to be different between CSHCN vs. non-CSHCN (p = 0.60).

The proportion of discrepancies found in CSHCN (56/2620, 2.1%) was not significantly different than the proportion found in non- CSHCN (142/5690, 2.5%, z = 1.1, p = 0.88). The association between CSHCN status and radiologic discrepancy was not significant (X2 = 0.990 df = 1, p = 0.32), and the distribution of the type of discrepancy was also not significant (X2 = 1.967 df = 3, p = 0.60). Additionally, PED census and imaging study time of day were not associated with differ- ences in discrepancy rates (rpb = 0.003, p = 0.76).

Table 1

Clinical features and medical equipment characterizing CSHCN.

chronic illness organ system (example diseases)

n

Pulmonary (asthma, cystic fibrosis)

1251

Neurological (epilepsy, neurofibromatosis)

696

Developmental (autism, intellectual disability)

552

Cardiac (cyanotic heart disease, VSD)

460

Gastrointestinal (short gut, Crohn’s)

399

Genetic (syndromes, chromosomal abnormalities)

394

Hematologic (Sickle cell disease, G6PD)

333

Oncologic (leukemia, solid tumors)

256

Endocrine (diabetes, panhypopituitarism)

161

Nephrologic (renal failure, Nephrotic syndrome)

156

Rheumatologic (lupus, idiopathic arthritis)

92

Transplant (solid organ, bone marrow)

58

Hepatic (Liver failure, biliary atresia)

55

Immunodeficiency (HIV, SCID, CGD)

44

Metabolic (Urea cycle, cyclic vomiting)

34

Medical equipment

n

Gastrostomy tube

394

Ventriculoperitoneal shunt

140

Tracheostomy tube

130

Gastrojejunostomy tube

58

Ventilator dependence

47

Central venous catheter (exposed)

44

Ileostomy/colostomy

36

Central venous catheter (burrowed)

34

Pacemaker

15

Vagal nerve stimulator

6

Hemodialysis catheter

4

Automated Implanted Defibrillator

2

Peritoneal dialysis catheter

2

Artificial urinary conduit (e.g. Mitrofanoff)

1

Glycopyrrolate (Baclofen) pump

1

Other

100

Table 2a provides a description for all high consequence [6] radiolog- ic discrepancies. All moderate consequence (36) radiologic discrepan- cies are presented as an appendix. No deaths resulted from a radiologic discrepancy.

Secondary Questions.

CHSCN presented with different chief complaints than non-CHSCN, as described in Table 3. In particular, trauma complaints were more like- ly to be found in non-CSHCN patients than CSHCN, and pulmonary com- plaints were more likely in CSHCN patients than non-CSHCN. This translated into differences found in types of radiologic images per- formed, where CSHCN had greater chest radiographs and non-CSHCN had greater extremity radiographs. Table 4 further describes the types of radiologic studies performed for CSHCN and non-CSHCN.

Discussion

Among 8310 visits, our overall radiologic discrepancy rate of 2.4% was similar to what is reported in the literature [4-7]. In a large adult study in a similar tertiary-care, urban setting, Petinaux et al. reported a radiologic discrepancy rate of 3.5% [11]. With a classification system similar to ours, they noted that emergency physicians have infrequently missed clinically significant (moderate consequence) findings and rare- ly missed emergent (high consequence) findings [11]. Our study

similarly demonstrated a low rate of clinically significant findings and rare high severity clinical consequences. However, one can’t underscore the impact of high severity clinical consequences, such as the finding of a large aortic aneurysm in adults [11] or findings suggestive of abuse in children. It is also interesting to mention that similar to the adult setting, high consequence findings were discovered on films being obtained for other reasons (i.e. chest radiograph being obtained to evaluate a pneu- monia discovering a non-accidental rib fracture in a pediatric patient versus a pelvic radiograph following a motor vehicle collision noting a large aortic aneurysm in an adult patient) [11]. These specific clinical scenarios emphasize the importance of taking a systematic approach to reviewing films, regardless of one’s training background [12-13].

Similar to the adult study by Petinaux et al. [11] and pediatric studies by Walsh-Kelly et al. [14-15], false-negative discrepancies were more prevalent than false-positive discrepancies, and false-negative Missed fractures accounted for the greatest number of moderate to high conse- quence radiologic discrepancies. Walsh-Kelly [15] found that false- negative chest radiograph interpretations and false-positive discrepan- cies had limited clinical impact. With regard to false-positive discrepan- cies, this may be related to the PEM practitioner treating based on clinical findings, which may be considered “overcalling” from a radiologist’s perspective [11]. While it is possible that false-positive dis- crepancies may result in admission and increasED resource utilization, particularly in CSHCN, this study was not able to address this aspect. Though as mentioned above, the existing literature suggests that false- positive discrepancies have limited clinical impact [15].

Ours is the first study to describe radiologic discrepancies in CSHCN. CSHCN represent a subset of pediatrics with greater medical complexity and thus greater potential for errors in their medical management [16-17]. Hospitalizations for CSHCN have doubled between 1991 and 1993 and 2003-2005, and CSHCN experience repeated hospitalizations, prolonged hospital stays, greater readmission rates, and greater inpa- tient Medical errors that non-CSHCN [17-18]. These have been attribut- ed to their underlying medical complexity and technological equipment dependence [17,19]. Malfunctions of their technological equipment alone contribute to their increased ED utilization and Hospital admission rates [17,19]. Their greater medical resource utilization also results in greater radiologic imaging utilization.

However, in contrast to these metrics demonstrating greater medi- cal complexity, greater resource utilization, and greater inpatient error rates [20], our study did not show greater radiologic discrepancy rates on imaging performed in the ED. Our findings may be explained by 2 factors about CSHCN. First, CSHCN are non-mobile and less likely to have traumatic injuries warranting extremity imaging, as shown by the differences in chief complaints within our data [21]. Extremity im- ages had the highest rate of discrepant reads. Second, it is possible that CSHCN – by virtue of their medical complexity – increased the clin- ical and radiologic concern for both the radiologist and ED practitioner, effectively over-treating based on a possible finding on a study. This is consistent with the literature on “overcalling” in radiologic discrepan- cies, where fractures and air-space disease account for the larger num- ber in this category, again reflecting the EM provider treating based on clinical findings [11].

An important but somewhat unexpected result is the fact that CSHCN did not demonstrate greater severity in clinical consequences due to radiologic discrepancies. Because ED radiologic discrepancies

Table 2

clinical actions following radiographic discrepancies in CHSCN and non-CHSCN.

(Fisher’s exact p = 0.60)

CSHCN (%)

Non-CSHCN (%)

Total

No consequence (incidental finding, or patient was already treated or admitted to the hospital)

27 (48.2%)

71 (50.0%)

98

Low consequence (patient was called for a pre-scription or follow-up with primary or specialty care)

20 (34.5%)

38 (26.8%)

58

Moderate consequence (patient was called to return to the PED for further treatment or repeat testing)

8 (14.3%)

28 (19.7%)

36

High consequence (patient was called to return for an admission or discharge to child protective services)

1 (1.8%)

5 (3.5%)

6

56

142

198

Table 2a

Descriptions of 6 high consequence discrepancies. PEM – pediatric emergency medicine attending. RadR -resident radiologist. RadA -attending radiologist. DCFS – Department of Child & Family Services (child protective services). n/a – no read available.

Age (years)

CSHCN

Chief complaint

radiological study/-ies

Discrepancy

Outcome

b 1

Trauma

CT: head

PEM-negative RadR-negative

Called back for skeletal survey and DCFS report; discharged home to parents

12

Infectious,

XR: tibia-fibula,

RadA-Skull fracture

PEM-no fracture

Admit, MRI;

b 1

Trauma

Respiratory,

knee, femur

XR: tibia-fibula

RadR-no fracture

RadA-Femoral condyle defect suggestive of osteomyelitis PEM-no fracture

Diagnosis of osteomyelitis

Called back for repeat images & skeletal survey, DCFS report; second RadA read as

Trauma

RadR-no fracture

RadA-Metaphyseal corner

negative. Discharged home to parents

5

General,

XR: spine

fracture

PEM-negative

Admit, cervical collar; bone biopsy revealed diagnosis of

1

Neurological

General

XR: femur,

RadR-n/a

RadA-C5 vertebra plana & soft tissue swelling

PEM-corner fracture

Langerhans Cell Histiocytosis

Called back for repeat images and DCFS discharge; eventually reunited. Child developed

tibia-fibula

RadR-no fracture

RadA-corner fracture and

progressive hypotonia and global developmental delay over 2 years

3

Autism

GI

XR: abdomen

wormian bones

PEM-high Stool burden

Admitted for bowel cleanout; appendicitis diagnosis not pursued

RadR-n/a

RadA-fecalith in right lower quadrant

for admitted patients were effectively managed by the inpatient team and therefore excluded from this study, the higher admission rates for CSHCN could result in fewer discrepant radiologic reads included for CSHCN. In other words, it is possible that an actual difference in discrep- ancy exists, but is masked by the higher admission rate among CSHCN.

Indeed CSHCN presented with greater pulmonary chief complaints, reflecting their greater propensity to have Respiratory infections [16], which is a more common cause for their hospital admission [19]. Alter- natively, a tertiary care setting familiar with the medical complexity of CSHCN and with the potential risk for diagnostic errors in caring for

Table 3

Significant associations between CSHCN status and chief complaints. Shaded cells indicate significantly higher percentage.

Chief complaint

(N = 8310 PED visits)

CSHCN (% of all chief complaints)

Non-CSHCN (% of

all chief complaints)

p-Value

General (e.g. weight loss, fatigue, not acting normal)

271 (10.3%)

543 (10%)

0.27

Cardiac (e.g. chest pain, syncope, palpitations)

130 (5%)

111 (2%)

<0.001

Dermatological (e.g. rash, pruritis)

23 (1%)

58 (1%)

0.54

Developmental (e.g. behavioral change, aggression)

4 (0.2%)

0

0.04

HEENT (e.g. pain, redness, discharge)

50 (2%)

90 (2%)

0.29

Gastrointestinal (e.g. abdominal pain, vomiting, jaundice)

668 (25%)

1637 (29%)

0.002

Genitourinary (e.g. bleeding, irregular period, testicular pain)

59 (2%)

180 (3%)

0.02

Hematological (e.g. vaso-occlusive pain, pallor, bruising)

35 (1%)

4 (0.1%)

<0.001

Infectious (e.g. fever)

586 (22%)

936 (16%)

<0.001

Nephrology (e.g. anuria, hematuria, oliguria)

18 (1%)

40 (1%)

0.93

Neurological (e.g. seizure, altered, headache, weakness)

180 (7%)

218 (4%)

<0.001

Pulmonary (e.g. cough, respiratory distress, apnea)

734 (28%)

889 (16%)

<0.001

Rheumatologic (e.g. joint pain)

23 (1%)

42 (1%)

0.50

Trauma

452 (17%)

2113 (37%)

<0.001

Table 4

Description of radiologic studies performed in CSHCN and non-CSHCN.

Radiologic study CSHCN (%

within total)

Non-CSHCN (%

within total)

staffed with pediatric emergency medicine physicians, the results may not be reflective of adult or community EDs that do not see the same proportion of CSHCN and non-CSHCN nor with pediatric radiologists on staff. However, the types and rates of discrepancies are similar to what is reported in the literature, further increasing the likelihood

Radiograph CORE (chest, abdomen, G- or GJ-tube

studies)

(67.6%)

Radiograph extremity

451

2042 (35.9%)

(17.2%)

XR chest

1235

1410 (24.8%)

(47.1%)

XR abdomen

524

1116 (19.6%)

(20.0%)

XR neck (soft tissue)

11 (0.4%)

50 (0.9%)

XR spine (C, T, L/S)

33 (1.3%)

83 (1.5%)

XR skeletal survey

4 (0.2%)

39 (0.7%)

CT brain

190 (7.3%)

313 (5.5%)

CT abdomen/pelvis

23 (0.9%)

39 (0.7%)

US abdomen

48 (1.8%)

100 (1.8%)

US pylorus

6 (0.2%)

94 (1.7%)

US pelvis

45 (1.7%)

144 (2.5%)

Other radiograph

237 (9.0%)

791 (13.9%)

1772

2456 (43.2%)

that our results are generalizable.

In our study, high severity clinical consequences from radiologic dis- crepancies were not related to ED census or time of day the imaging study was performed. Some studies have shown greater error rates in the final two hours of a radiology resident shift while other studies have shown similar rates regardless of time [23-25]. Although our data did not find a relationship with ED census or time of day, the low number of discrepancies, especially with clinical consequences, may be insufficient to reveal an effect.

A single research assistant was responsible for abstracting the radio- logic data. The research assistant was extensively trained and all physi- cian investigators were available to provide clarification, with any remaining questions reconciled by consensus thus limiting errors in coding. Specific radiologic discrepancies and medical records were reviewed by PED physician investigators (AF, KK, DL) and confirmed by a pediatric radiologist investigator (HL).

CSHCN, may be more risk averse than other settings and over-treat, thus lowering the risk for false negative reads and clinically consequential discrepancies. This would support CSHCN being primarily treated in ter- tiary care settings [22]. Furthermore, repeat visits to a tertiary care cen- ter allows all practitioners access to prior imaging for comparison and potentially improved accuracy of reads.

Limitations

As with any retrospective study design, the accuracy and complete- ness of the data is limited by what is contained within the medical re- cord. As a study within a single, freestanding children’s hospital

Conclusion

Though the frequency and type of radiologic studies performed were different between children with and without special Health care needs, we found no significant difference in the rate of radiologic discrepancies nor the rate of clinically significant radiologic discrepancies between children with and without special health care needs.

Acknowledgments

We would like to thank Ms. Phung K Pham for her data analysis and statistical assistance.

Appendix A

Descriptions of 36 moderate consequence discrepancies. PEM – pediatric emergency medicine attending. RadR -resident radiologist. RadA -attending radiologist. DCFS – Department of Child & Family Services (child protective services). n/a – no read available.

Age (years)

CSHCN

Chief complaint

Radiological study/-ies

Discrepancy

Outcome

9

4

Asthma

Asthma

Respiratory

Respiratory

XR: chest

XR: chest

PEM-negative RadR-negative

RadA-lingular pneumonia

PEM-negative

Return for further Imaging

Return for further

16

Asthma

Genito-urinary

US: scrotum

RadR-negative

RadA-mediastinal fullness PEM-mild decreasED flow

laboratories (rule out leukemia)

Return for further Imaging

6

Asthma

Respiratory

XR: chest

RadR-unchanged from before RadA-cannot rule out torsion

PEM-negative

Return for further Imaging

6

Asthma

Trauma

XR: spine

RadR-negative RadA-pneumonia

PEM-no fracture

Return for further Imaging

16

Asthma

General

XR: hip

RadR-no fracture

RadA-calcification suggestive of mass PEM-no slipped capital femoral epiphysis

Return for further Imaging

6

Chromosomal Anomaly, Absent Corpus Callosum,

Genito-urinary

US: scrotum

RadR-n/a

RadA-paucity of abdominal gas concerning for intussusception PEM-negative

Return for Urological

Chronic Lung Disease, Central Sleep apnea

RadR-no venous flow but good arterial

flow

consultation

11

Known cardiac murmur

Trauma

XR: ankle

RadA- peri-testicular dysmorphic mass

PEM-negative

Return for a procedure

7

Trauma

XR: wrist

RadR-n/a

RadA-occult fracture PEM-negative

Return for a procedure

RadR-negative

(continued)

Age (years)

CSHCN

Chief complaint

Radiological study/-ies

Discrepancy

Outcome

17

Trauma

XR: foot

RadA-possible fracture PEM-no fracture

Return for further

b 1

Infectious

XR: chest

RadR-lucency noted in phalanx RadA-phalangeal fracture

PEM-no infiltrate

Imaging/procedutre

Return for ED evaluation

b 1

GI

XR: abdomen,

RadR-no infiltrate

RadA-concern for cardiomegaly

PEM-no intussusception/pyloric stenosis

Return for IV

4

GI,

US: pylorus, US: abdomen

XR: hand

RadR-no intussusception/pyloric stenosis

RadA-debris in bladder, consider UTI PEM-no fracture

hydration, urine testing

Return for a procedure

trauma

RadR-no fracture

8

Trauma

XR: wrist, elbow

RadA-phalangeal fracture

PEM-elbow effusion

Return for further Imaging

2

GI

XR: abdomen

RadR-elbow effusion

RadA-effusion & Salter-Harris II radial fracture

PEM-negative

Return for further Imaging

RadR-negative

RadA-abnormal bowel gas pattern

b 1

2

Trauma

General

XR: elbow, humerus, clavicle

XR: chest,

PEM-no fracture RadR-no fracture

RadA-concern for occult supracondylar fracture

PEM-n/a

Return for further Imaging

Return for further Imaging

9

Trauma

abdomen

XR: finger

RadR-density near T6 vertebra RadA-thoracic rib lytic lesion

PEM-no fracture

Return for a procedure

2

General,

XR: abdomen

RadR-no fracture

RadA-phalangeal fracture PEM-negative

Return for further Imaging

2

GI

GI

XR: abdomen

RadR-negative

RadA-paucity of gas, rule out Abdominal mass

PEM-negative

Return for a procedure

16

Trauma

CT: head

RadR-negative

RadA-paucity of gas, rule out abdominal mass

PEM-negative

Return for ED evaluation

6

Genito-urinary

US: scrotum

RadR-negative

RadA-noted a pineal calcification, possi- ble brain mass

PEM-no torsion

Return for ED evaluation

12

Trauma

XR: finger

RadR-no torsion

RadA-no torsion; symptoms may be too early; repeat US recommended

PEM-reduced fracture

Return for a procedure

12

GI

XR: abdomen

RadR-reduced fracture

RadA-Different phalangeal fracture PEM-negative

Return for further Imaging

14

Trauma

XR: knee

RadR-n/a

RadA-appendicolith PEM-patellar fraying

Return for further Imaging

b 1

Trauma

XR: forearm

RadR-patellar fraying

RadA-medial condyle irregularity PEM-no fracture

Return for a procedure

RadR-no fracture

RadA-avulsion fracture

14

1

Trauma

Trauma

XR: knee

XR: hand, wrist,

PEM-negative RadR-negative

RadA-high patella, concern for tendon rupture

PEM-no fracture

Return for further Imaging/procedure

Return for further Imaging

forearm

RadR-no fracture

RadA-cortical undulation concerning for buckle fracture

3

2

GI

Trauma

XR: abdomen

XR: elbow

PEM-high stool burden RadR-high stool burden

RadA-left lower quadrant paucity of gas, concern for abdominal mass

PEM-negative

Return for further Imaging

Return for a procedure

RadR-negative

RadA-anterior fat pad concerning for supracondylar fracture

(continued on next page)

(continued)

Age (years)

CSHCN

Chief complaint

Radiological study/-ies

Discrepancy

Outcome

16

12

Trauma

Trauma

XR: knee

XR: foot, ankle

PEM-Osgood-Schlatter disease RadR-no fracture

RadA-proximal tibial fracture

PEM-no fracture

Return for a procedure

Return for ED evaluation

7

Infectious,

XR: chest,

RadR-n/a

RadA-Distal fibular lucency PEM-high stool burden

Return for further Imaging

b 1

neurological, respiratory

Trauma

abdomen

XR: knee

RadR-high stool burden RadA-appendicolith

PEM-Patellar fracture

Return for Orthopedic

RadR-n/a

RadA-tibial spine fracture

consultation

3

b 1

Trauma

Respiratory

XR: elbow

XR: neck soft

PEM-no fracture RadR-no fracture

RadA-humeral neck fracture

PEM-no foreign body

Return for a procedure

Return for further Imaging

tissue

RadR-no foreign body

RadA- tracheal deviation. Consider medi- astinal mass

References

  1. Lufkin KC, Smith SW, Matticks CA, Brunette DD. Radiologists’ review of radiographs interpreted confidently by emergency physicians infrequently leads to changes in patient management. Ann Emerg Med 1998;31(2):202-7. https://doi.org/10.1016/ S0196-0644(98)70307-5.
  2. Chung JH, Strigel RM, Chew AR, Albrecht E, Gunn ML. Overnight resident interpreta- tion of torso CT at a level 1 trauma center. An Analysis and Review of the Literature. Acad Radiol 2009;16(9):1155-60. https://doi.org/10.1016/j.acra.2009.02.017.
  3. Erly WK, Berger WG, Krupinski E, Seeger JF, Guisto JA. Radiology resident evaluation of head CT scan orders in the emergency department. Am J Neuroradiol 2002;23(1): 103-7.
  4. Niska R, Bhuiya F, Xu J. National hospital ambulatory medical care survey: 2007 emergency department summary. Natl Health Stat Rep 2010;26:1-31http://www. ncbi.nlm.nih.gov/pubmed/20726217.
  5. Friedman SM, Merman E, Chopra A. Clinical impact of diagnostic imaging discrepan- cy by radiology trainees in an urban teaching hospital emergency department. Int J Emerg Med 2013;6(1):24. https://doi.org/10.1186/1865-1380-6-24.
  6. Shirm SW, Graham CJ, Seibert JJ, Fiser D, Scholle SH, Dick RM. Clinical effect of a qual- ity assurance system for radiographs in a pediatric emergency department. Pediatr Emerg Care 1995;11(6):351-4. https://doi.org/10.1097/00006565-199512000- 00005.
  7. Soudack M, Raviv-Zilka L, Ben-Shlush A, Jacobson JM, Benacon M, Augarten A. Who should be reading chest radiographs in the pediatric emergency depart- ment? Pediatr Emerg Care 2012;28(10):1052-4. https://doi.org/10.1097/PEC. 0b013e31826caf3f.
  8. Newacheck PW, Strickland B, Shonkoff JP, et al. An epidemiologic profile of children with special health care needs. Pediatrics 1998;102(1 Pt 1):117-23. https://doi.org/ 10.1542/peds.102.1.117.
  9. Mcpherson M, Arango P, Fox H, et al. A new definition of children with special health care needs. Pediatrics 1998;102(1):137-9. https://doi.org/10.1007/978-1-4613- 2991-6.
  10. Van der Lee JH, Mokkink LB, Grootenhuis MA, Heymans HSOM. Definitions and mea- surement of chronic health conditions in childhood: a systematic review. JAMA 2007;297(24):2741-51. https://doi.org/10.1001/jama.297.24.2741.
  11. Petinaux B, Bhat R, Boniface K, Aristizabal J. Accuracy of radiographic readings in the emergency department. Am J Emerg Med 2011;29(1):18-25. https://doi.org/10. 1016/j.ajem.2009.07.011.
  12. Robinson PJ. Radiology’s achilles’ heel: error and variation in the interpretation of the Rontgen image. Br J Radiol 1997;70(839):1085-98.
  13. Peterson C. factors associated with success or failure in radiological interpretation: diagnostic thinking approaches. Med Educ 1999;33(4):251-9. https://doi.org/10. 1046/j.1365-2923.1999.00295.x.
  14. Walsh-Kelly CM, Hennes HM, Melzer-Lange MD. False-positive preliminary radio- graph interpretations in a pediatric emergency department: clinical and economic impact. Am J Emerg Med 1997;15(4):354-6.
  15. Walsh-Kelly CM, Melzer-Lange MD, Hennes HM, et al. Clinical impact of radiograph misinterpretation in a pediatric ED and the effect of physician training level. Am J Emerg Med 1995;13(3):262-4. https://doi.org/10.1016/0735-6757(95)90196-5.
  16. Sacchetti A, Sacchetti C, Carraccio C, Gerardi M. The potential for errors in children with special health care needs. Acad Emerg Med 2000;7:1330-3. https://doi.org/ 10.1111/j.1553-2712.2000.tb00484.x.
  17. Coller RJ, Lerner CF, Eickhoff JC, et al. Medical complexity among children with spe- cial health care needs: a two-dimensional view. Health Serv Res 2016;51(4): 1644-69. https://doi.org/10.1111/1475-6773.12416.
  18. Slonim AD, Lafleur BJ, Ahmed W, Joseph JG. Hospital-reported Medical errors in chil- dren. Pediatrics 2003;111(3):617-21. https://doi.org/10.1542/peds.111.3.617.
  19. Russell CJ, Simon TD. Care of children with medical complexity in the hospital setting. Pediatr Ann 2014;43(7):e157-2. https://doi.org/10.3928/00904481- 20140619-09.
  20. Sacchetti A, Sacchetti C, Carraccio C, Gerardi M. The potential for errors in children with special health care needs. Acad Emerg Med 2000;7:1330-3. https://doi.org/ 10.1111/j.1553-2712.2000.tb00484.x.
  21. Bramlett MD, Read D, Bethell C, Blumberg SJ. Differentiating subgroups of children with special health care needs by health status and complexity of health care needs. Matern Child Health J 2009;13(2):151-63. https://doi.org/10.1007/s10995- 008-0339-z.
  22. Auerbach M, Whitfill T, Gawel M, et al. Differences in the quality of pediatric resus- citative care across a Spectrum of emergency departments. JAMA Pediatr 2016; 170(10):987-94. https://doi.org/10.1001/jamapediatrics.2016.1550.
  23. Davenport MS, Ellis JH, Khalatbari SH, Myles JD, Klein KA. Effect of work hours, case- load, shift type, and experience on resident call performance. Acad Radiol 2010; 17(7):921-7. https://doi.org/10.1016/j.acra.2010.03.006.
  24. Ruutiainen AT, Scanlon MH, Itri JN. Identifying benchmarks for discrepancy rates in preliminary interpretations provided by radiology trainees at an academic institu- tion. J Am Coll Radiol 2011;8(9):644-8. https://doi.org/10.1016/j.jacr.2011.04.003.
  25. Shah NA, Hoch M, Willis A, Betts B, Patel HK, Hershey BL. Correlation among on-call resident study volume, discrepancy rate, and turnaround time. Acad Radiol 2010; 17(9):1190-4. https://doi.org/10.1016/j.acra.2010.06.003.

Leave a Reply

Your email address will not be published. Required fields are marked *