|Year : 2020 | Volume
| Issue : 2 | Page : 116-121
Accuracy of diagnoses of skin diseases in a pediatric emergency room: A retrospective study
Suzan A AlKhater
Department of Pediatrics, College of Medicine, Imam Abdulrahman Bin Faisal University, Dammam; Department of Pediatrics, King Fahad University Hospital, Al-Khobar, Saudi Arabia
|Date of Submission||14-Nov-2019|
|Date of Acceptance||11-Dec-2019|
|Date of Web Publication||10-Nov-2020|
Dr. Suzan A AlKhater
Department of Pediatrics, College of Medicine, Imam Abdulrahman Bin Faisal University, Dammam; King Fahad University Hospital, P.O. Box: 2208, Al-Khobar 31952
Source of Support: None, Conflict of Interest: None
Background: Cutaneous disorders are common in children. Few studies have assessed the diagnostic accuracy of pediatricians in diagnosing skin disorders. Purpose: In this study, the accuracy of the diagnoses of skin diseases made in a pediatric emergency room (ER) was determined. Methods: Data were obtained from a retrospective study of children with dermatological disorders seeking emergency services in the pediatric ER in a teaching hospital in Saudi Arabia. Descriptive statistics were used to determine the distribution of patients and time characteristics. Crude and adjusted logistic regression models, controlling for demographic and time-associated characteristics, were used to determine the association between clinical and other characteristics and diagnostic accuracy of pediatricians. Results: There were statistically significant differences in the overall distribution of the study population by gender, season, time of week, shift, and referrals (P < 0.05). After adjusting for these factors, every unit increase in age (years) was associated with a 38% increase (odds ratio: 1.38; 95% confidence interval: 1.18–1.61) in the odds of pediatricians being likely to have the correct diagnosis. There were significant differences between the diagnosis made by pediatrician and the final diagnoses made by senior staff (P < 0.001). Infections and neonatal skin diseases were underdiagnosed, while overdiagnoses of allergic and hypersensitivity reactions were observed. Conclusion: The results emphasize the importance of orienting pediatric trainees on the dermatological aspects of diseases, particularly on the topics involving infectious, allergic and hypersensitivity reactions, and neonatal skin disorders. This points out the importance of making dermatology rotation as a core requirement during the residency training rather than just optional electives.
Keywords: Children, curriculum, dermatology, diagnosis, pediatricians, training
|How to cite this article:|
AlKhater SA. Accuracy of diagnoses of skin diseases in a pediatric emergency room: A retrospective study. J Dermatol Dermatol Surg 2020;24:116-21
| Introduction|| |
Cutaneous disorders are commonly encountered in the pediatric age group. Depending on the urgency of the complaint, patients seek outpatient dermatology clinics or emergency service. From a pediatrician's perspective, emergency room (ER) visits related to dermatological complaints have recently received increased attention because they are perceived as a burden on an already busy service.,
An estimated 4%–16% of the total visits in pediatric emergency units are related to the disorders of skin and appendages.,, In a previous report from our institute, this frequency was approximately 4.7% of total ER visits, equivalent to an average of 5.7 patients per day. Although not exceptionally high, the cases observed were of a wide spectrum, ranging from mild to the more severe cases.,, This issue results in a daunting diagnostic process for pediatricians as often little or no formal dermatology training is received during the pediatric residency training years., This pattern of training of pediatric residents is observed throughout the training institutes of the Kingdom of Saudi Arabia. Furthermore, residents' academic day activities, intended to provide lectures dedicated to certain topics in their field of specialty, do not regularly include teaching on dermatology topics.
During the Saudi pediatric residency program, formal dermatology training is usually received during elective rotations that are optional and do not usually exceed 4 weeks, in total, during the 4 years of training. However, there is a general lack of published studies in Saudi Arabia in regard to the effectiveness of such teaching methods. Previous studies from Swiss, France, Ireland, Mexico, and the US have reported significant deficiencies in diagnosing dermatological disorders among pediatricians, due to a general lack of dermatology teaching in the pediatric residency training.,,,, This may result in the inability to handle the increasing demands in the field.
Pediatric residents need to be comfortable and competent when dealing with dermatological complaints in their field, as they may likely face a wide spectrum of disorders. Indeed, many pediatric doctors reported difficulties in dealing with dermatology cases, and 25% indicated the need to consult dermatology service for patients with skin complaints. Furthermore, pediatricians missed more than half the cases related to dermatological disorders of an infectious etiology, the most common disorder encountered in the reported study cohort. In addition, in one report, the diagnoses made by pediatricians on dermatological cases were modified by a dermatologist in 42% of cases.
The aim of this study was to assess the diagnostic accuracy of pediatric trainees and specialists in diagnosing skin-related disorders and to provide suggestions for improvement. Furthermore, possible contributing factors were evaluated, including whether a busy ER service may impact the accuracy to diagnose skin-related complaints.
| Methods|| |
Data source and study population
Data were obtained from a retrospective study performed during 2014 and involves all children with dermatological disorders seeking emergency services in the pediatric ER of a university teaching hospital in AL-Khobar city, Saudi Arabia. This hospital is considered a large referral center that provides tertiary care and receives over 40,000 visits per year in the pediatric ER alone. The pediatric ER, a division of the pediatric medical department, is responsible for all children visiting the ER for medical complaints, including skin-related issues and is staffed by on-site pediatric specialists and trainees. Patients with dermatological complaints are evaluated by the pediatric ER staff. Referral to specialized care is usually performed in urgent cases. As dermatology cases are mostly nonurgent, referrals are usually made to ambulatory dermatology outpatient care, whereas on-call dermatologists are usually consulted in severe cases. The study received ethical approval from the Institutional Review Board in accordance with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.
All pediatric ER visitors aged ≤13 years in 2014, with a primary and final diagnosis involving a disorder affecting the skin and its appendages, the hair and nails, were retrospectively reviewed. There were 2,070 dermatological cases identified, representing over 80 different types of dermatological disorders. To ensure an equal distribution of patients across all seasons, the analysis included patients who visited ER in the 1st week of each month from January 1, 2014, to December 30, 2014. Patients with trauma, bites and stings, foreign body, or burns were excluded as the diagnosis in such cases was self-explanatory. A detailed chart review was made for all the selected cases. Patients that are known to the dermatology service were excluded, and only new patients were included in the study. The final sample size was n = 442. The following data were obtained: age, gender, time and date of the visit, type of referral (self-referred or physician referral), manner of follow-up, including whether the patient was referred to dermatology or other services, and the pediatric ER diagnosis (PED), which is given by the pediatric residents or pediatric specialists covering ER at the time of the visit. In addition, cases with dermatology referrals or consultations were also reviewed.
The final diagnosis was defined as the diagnosis made by a senior staff covering the ER at the time of the visit, during subsequent ER visits, or during the visits to the pediatric or dermatology specialty clinics. The senior staff were usually pediatric consultants or dermatology consultants. For patients not seen by a senior staff, laboratory investigations and details of the medical notes were used to identify the final diagnosis. The diagnosis was considered “unknown” if no diagnosis was reached because the medical notes were inconclusive or the patient was not evaluated by senior staff or referred to specialized care.
The diagnostic accuracy was operationalized as a binary variable by assessing the diagnostic accuracy of PED compared to the final diagnosis. The variable was coded as 1 if PED matched the final diagnosis and 0 if it did not. Dermatological disorders were grouped as follows: (1) infections, (2) allergic disorders and hypersensitivity reactions, (3) systemic and inflammatory disorders and purpura, (4) neonatal dermatological complaints, (5) miscellaneous, (6) nonspecific maculopapular rash, and (7) unknown/unidentified skin disorder.
Referrals were operationalized as at least one visit to the dermatology clinic, if the patient was seen by the dermatologist on-call in the ER on the date of the visit, or when a documented referral was given to the patient for the outpatient dermatology clinics.
Descriptive statistics were used to determine the distribution of patient and time-associated characteristics and referrals overall and by diagnostic accuracy of pediatricians. PEDs and final diagnoses were also compared to determine if there were statistically significant differences (P < 0.05). Moreover, PEDs and final diagnoses were cross-tabulated to determine the agreement of diagnoses of separate categories (and a lack thereof) of diagnostic codes. Crude and adjusted logistic regression models were used to determine the association between clinical and other characteristics and the diagnostic accuracy of the pediatric clinicians. Adjusted models controlled for all demographic, clinical, and time-associated characteristics. The total number of dermatological cases seen and the diagnostic accuracy of the pediatric clinicians were also graphed by the month and day of the week of the diagnosis. This comparison was performed to determine if the diagnostic accuracy of a pediatric clinician is affected by the crowdedness of the ER, which varied by the specific days of the week or seasons of the year. Compared to other seasons or shifts, the winter season and afternoon shifts generally tend to be busier times. This difference may result in inaccuracies in diagnosing the dermatological conditions. All analyses were conducted in SAS version 9.4 (SAS Institute, Cary, NC).
| Results|| |
Overall, pediatric ER clinicians were incorrect in their diagnosis in over one-quarter of cases (26.7%; n = 119), and approximately one in eight were diagnosed as unknown (13.8%, n = 61). [Table 1] shows the distribution of patients, time-associated characteristics, and referrals overall and by the pediatrician's diagnostic accuracy. There were significant differences in the overall distribution of the study population by gender, season, time of week, shift, and referrals (P < 0.05). The majority of the cases were male (55%) observed in spring (28%) and winter (27%) and during the week from Sunday to Thursday (67%). Approximately 48% of cases were seen in the afternoon shift (3–11 PM), and only 10% were given a referral. Significant differences in season and referrals were seen by the diagnostic accuracy. A lower percentage of cases were accurately diagnosed in spring compared to those that were inaccurately diagnosed (25% vs. 38%). However, compared to the proportion of inaccurately diagnosed cases, higher proportion of cases was accurately diagnosed in summer (28% vs. 20%) and winter (28% vs. 24%).
|Table 1: Distribution of patient and time characteristics and referrals overall and by the pediatrician's diagnostic accuracy|
Click here to view
There were statistically significant differences between PED and the final diagnoses (P < 0.001) [Table 2]. There was an underdiagnosis of infections, neonatal dermatological complaints, and nonspecific maculopapular rash. An example of an underdiagnosed infectious case was of an 11-year-old female who was diagnosed by a pediatric doctor to have dermatitis artefacta, based on the findings of multiple hyperpigmented lesions in the face and genitalia. Accordingly, she was labeled a victim of child abuse. However, the final diagnosis of the dermatologists was bullous impetigo. On the other hand, there was overdiagnosis of allergic disorders and hypersensitivity reactions and slight overdiagnosis of systemic and inflammatory disorders and purpura. In that regard, a 3-year-old female with the diagnosis of urticaria, made by a pediatrician upon the finding of an extensive skin rash, was subsequently diagnosed, by the dermatology consultant, as port-wine stain that involved the left shoulder, left side of the neck, and left arm and forearm. Aside from that, the diagnoses of miscellaneous conditions were the same for PED and the final diagnosis. However, conditions were almost three times more likely to be indicated as “unknown” by pediatric trainees than by dermatologists and the pediatric consultant (final diagnosis) (P ≤ 0.001).
|Table 2: Comparison of pediatric emergency room diagnoses and final diagnoses of experts|
Click here to view
[Table 3] shows the cross tabulation of the final diagnosis and PED. Approximately 86% of infections, 83% of allergic disorders and hypersensitivity reactions, 77% of systemic and inflammatory disorders and purpura, 29% of neonatal dermatological complaints, 64% of all miscellaneous, and 9% of nonspecific maculopapular rash diagnoses, as diagnosed by the dermatologist were identified as such by a pediatrician. The majority of nonspecific maculopapular rashes were identified as allergic disorders and hypersensitivity reactions. The cross-tabulation also shows that even though the same number of cases was diagnosed as “unknown” by an expert and a pediatrician, as shown in [Table 2], some “unknowns” were misdiagnosed. These misdiagnosed cases included seven cases diagnosed with allergic disorders and hypersensitivity reactions and one diagnosed with miscellaneous.
|Table 3: Cross-tabulation of the final diagnosis and pediatric emergency room diagnosis|
Click here to view
[Table 4] shows the crude and adjusted associations between the patient and other characteristics and the pediatrician's diagnostic accuracy. After adjusting for gender, seasons, time of week, shift and referrals, every unit increase in age (years) was associated with a 38% increase (odds ratio [OR]: 1.38; 95% confidence interval [CI]: 1.18–1.61) in the odds of pediatric clinicians making the correct diagnosis. After adjusting for age, gender, time of week, shift and referrals, compared to other seasons, the spring season was associated with a 46% decrease (OR: 0.54; 95% CI: 0.30–0.97) in the odds of pediatricians making the correct diagnosis. After adjusting for age, gender, seasons, time of week, and shift, compared to no referral (being discharged without referral), referral was associated with a 74% decrease (OR: 0.26; 95% CI: 0.13–0.51) in the odds of pediatricians making the correct diagnosis. There were no significant associations between gender, time of the week, shift, and the pediatrician's diagnostic accuracy [Figure 1].
|Table 4: Association between characteristics and the pediatrician's diagnostic accuracy|
Click here to view
|Figure 1: Adjusted associations between patient and time characteristics and referrals and the pediatrician's diagnostic accuracy|
Click here to view
[Figure 2] shows the total number of dermatological cases seen and the diagnostic accuracy by month (A) and day of the week (B). Closer lines indicate greater diagnostic accuracy. Diagnostic accuracy was high in June and October but was lower in other months. Diagnostic accuracy was also better on Thursday than on other days of the week.
|Figure 2: (a) A total number of dermatological cases seen and diagnostic accuracy by month. (b) Total number of dermatological cases seen and diagnostic accuracy by day of the week|
Click here to view
| Discussion|| |
The intent of this study was to assess the diagnostic accuracy of pediatric trainees in diagnosing cutaneous disorders and to assess the contributing factors. The main findings of this study were that the diagnosis of pediatric trainees was less likely to be accurate in the spring season and when a referral was made compared to that in other seasons and when a referral was not made. There was also a tendency for pediatricians to underdiagnose infections and neonatal dermatological complaints and to overdiagnose allergic disorders and hypersensitivity reactions compared to other diagnoses. Approximately 14% of the cases did not receive any diagnosis.
Dermatological disorders include a broad spectrum of diseases. Pediatricians, internists, and primary care physicians have an expanding role in the patient care. There is a growing concern about the inadequacy of dermatological training among nondermatologists managing such patients., Dermatology training during residency is usually obtained through electives, which are optional. Despite being perceived by trainees as a privilege, the dermatology elective is available to only a small proportion of those trainees.
The current study showed a diminished accuracy of pediatric trainees to diagnose dermatological disorders. This finding supports previous research published on this topic that showed a general lack of the pediatrician's to accurately diagnose dermatological disorders.,,,, One-quarter of dermatological cases receive no diagnosis by pediatricians., In a study assessing the agreement between emergency pediatricians and dermatology consultants, diagnostic agreement was reached in only 58% of the cases, while treatment prescribed by a pediatrician was changed by dermatologists in 30% of the cases. Similar results were observed in a study that reviewed the utilization of dermatology consultations over a period of 36 months. The present study found that the diagnostic agreement between the pediatric emergency doctor and the dermatologist is 58% only, of a total 347 dermatological consultations seen during that period. Similar to the findings in this study, their study demonstrated that the highest rate of referrals were those with infectious diseases and allergies, mostly atopic dermatitis. The study suggested educating pediatricians on the more commonly encountered diseases seen in childhood skin-related disorders.
Proper collaboration between the pediatric and dermatology service could provide a learning opportunity for pediatricians and improve patient care and outcomes. A lack of such communication can result in misdiagnosis and may have serious consequences, even in milder cases. In fact, some of the patients seen in this study cohort have illustrated the importance of proper pediatric-dermatology communication. An example of this was the misdiagnosis of the 11-year-old female with bullous impetigo as dermatitis artefacta. She subsequently was mislabeled as a victim of child abuse. Interestingly, a similar case was reported in a previous study that assessed the diagnostic concordances between junior ER doctors and dermatologists. Furthermore, misdiagnosis of some of the cases had resulted in unnecessary treatment measures. This was noted in the finding of the 3-year-old female who was diagnosed by the pediatrician as a case of urticaria while she actually suffered from an extensive skin involvement with port-wine stain. The parents were instructed, unnecessarily, to avoid cow milk protein because it was thought to aggravate the condition. Better pediatric-dermatology communication could have prevented unnecessary actions and treatment measures. Furthermore, this study demonstrated that 14% of the cases diagnosed by the pediatric trainees were unknown, a proportion similar to a previous study that reported that 18% of dermatological cases did not receive any diagnosis.
To improve the communication between the pediatric ER clinicians and dermatologists, a dermatology coverage has been arranged in some tertiary centers., However, implementation of a structured pediatric emergency-dermatology coverage may not be justified, especially if the frequency of the visits is not very high. Nevertheless, a proper collaboration between the pediatric and the dermatology service should be emphasized. This collaboration can be enforced through organizing steady dermatology ER coverage by dermatology trainees, as this can facilitate the exchange of knowledge and skills related to diseases common to both their fields. Similarly, common training sessions can be arranged to facilitate better communication between these specialties that share many aspects. Alternatively, online modules can become effective alternatives to improve the knowledge of specific conditions, especially when there is a lack of adequate exposure to clinical scenarios.
In the current study, there was a tendency to underdiagnose certain disorders while others were overdiagnosed. In addition, a lack of adequate information provided by the ER doctor covering the service at the time of consultation contributed to the majority of unknown diagnoses made by the expert. These findings imply that additional training is needed for ER pediatricians in the diseases that are more commonly encountered in the ER setting. Furthermore, instructions to ER doctors to provide adequate information in the files are necessary in order to for the experts to reach a final diagnosis and reduce uncertainties in the diagnosis.
The findings of the current study must be considered within the context of limitations. The busy ER setting may not be the optimal place or time to assess pediatricians' diagnostic competencies. The need to see many patients in a short time can negatively influence the diagnostic accuracy of a pediatric doctor. Furthermore, as cases attending the ER are usually urgent, mild disorders are likely to be overlooked by the physicians who are naturally obliged to shorten the visit time for the cutaneous cases, which are usually perceived as mild, to dedicate more time to cases that are more severe in nature. This expectation may explain why approximately one in eight cases was diagnosed as “unknown” and a quarter of cases were misdiagnosed by the pediatrician.
Despite these limitations, the current study had some strengths. The current study identified obvious gaps in the diagnostic accuracy of pediatricians in the ER, which is probably caused by inadequate exposure to the dermatology field. Adjusted models controlled for all demographic, clinical, and time-associated characteristics. Therefore, the adjusted results accounted for characteristics that may have influenced the “true association” between each characteristic and the pediatrician's diagnostic accuracy. In addition, the sample size was large at n = 442 cases.
| Conclusion|| |
The current study identified deficiencies in the accuracy to diagnose cutaneous disorders in the pediatric ER. The results emphasize the importance of orienting pediatric trainees on the dermatological aspect of disease related to their field and making dermatology a requirement during their residency training rather than just an elective. Furthermore, proper pediatrics-dermatology communication is required to improve the quality of patient care provided.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
Chamlin SL. Cases in pediatric dermatology. Pediatr Ann 2013;42:18.
Landolt B, Staubli G, Lips U, Weibel L. Skin disorders encountered in a Swiss pediatric emergency department. Swiss Med Wkly 2013;143:w13731.
AlKhater SA, Dibo R, Al-Awam B. Prevalence and pattern of dermatological disorders in the pediatric emergency service. J Dermatol Dermatol Surg 2017;21:7-13.
Kramkimel N, Soussan V, Beauchet A, Duhamel A, Saiag P, Chevallier B, et al
. High frequency, diversity and severity of skin diseases in a paediatric emergency department. J Eur Acad Dermatol Venereol 2010;24:1468-75.
Wang E, Lim BL, Than KY. Dermatological conditions presenting at an emergency department in Singapore. Singapore Med J 2009;50:881-4.
Auvin S, Imiela A, Catteau B, Hue V, Martinot A. Paediatric skin disorders encountered in an emergency hospital facility: A prospective study. Acta Derm Venereol 2004;84:451-4.
Dolan OM, Bingham EA, Glasgow JF, Burrows D, Corbett JR. An audit of dermatology in a paediatric accident and emergency department. J Accid Emerg Med 1994;11:158-61.
Soriano-Hernández YL, Orozco-Covarrubias L, Tamayo-Sánchez L, Durán-McKinster C, Sosa-de-Martínez C, Ruiz-Maldonado R. Exanthems in hospitalized pediatric patients: Concordance between pediatric and dermatological diagnoses. Dermatology 2002;204:273-6.
Fleischer AB Jr., Herbert CR, Feldman SR, O'Brien F. Diagnosis of skin disease by nondermatologists. Am J Manag Care 2000;6:1149-56.
Drucker AM, Cavalcanti RB, Wong BM, Walsh SR. Teaching dermatology to internal medicine residents: Needs assessment survey and possible directions. J Cutan Med Surg 2013;17:39-45.
McCarthy GM, Lamb GC, Russell TJ, Young MJ. Primary care-based dermatology practice: Internists need more training. J Gen Intern Med 1991;6:52-6.
Khorsand K, Brandling-Bennett HA. Deficiencies in dermatologic training in pediatric residency: Perspective of pediatric residency program directors. Pediatr Dermatol 2015;32:819-24.
Julián-González RE, Vargas-de Julián JV, Valdebran-Canales M. Diagnostic accuracy among pediatricians and pediatric residents: A brief research study from El Salvador. Int J Dermatol 2016;55:e620-e621.
Moon AT, Castelo-Soccio L, Yan AC. Emergency department utilization of pediatric dermatology (PD) consultations. J Am Acad Dermatol 2016;74:1173-7.
Martínez-Martínez ML, Escario-Travesedo E, Rodríguez-Vázquez M, Azaña-Defez JM, Martín de Hijas-Santos MC, Juan-Pérez-García L. Dermatology consultations in an emergency department prior to establishment of emergency dermatology cover. Actas Dermosifiliogr 2011;102:39-47.
Grillo E, Vañó-Galván S, Jiménez-Gómez N, Ballester A, Muñoz-Zato E, Jaén P. Dermatologic emergencies: Descriptive analysis of 861 patients in a tertiary care teaching hospital. Actas Dermosifiliogr 2013;104:316-24.
Craddock MF, Blondin HM, Youssef MJ, Tollefson MM, Hill LF, Hanson JL, et al
. Online education improves pediatric residents' understanding of atopic dermatitis. Pediatr Dermatol 2018;35:64-9.
[Figure 1], [Figure 2]
[Table 1], [Table 2], [Table 3], [Table 4]