Guidelines for Diagnosis of Child Physical Abuse—Let’s Get Back to Work!

The World Health Organization considers child maltreatment an international priority and currently develops a vast program of child physical abuse (CPA) prevention.¹ Therefore, availability of optimal guidelines for the key actors involved in early detection of CPA is clearly identified as one of the major interventions to prevent child maltreatment. However, in order to be considered optimal, guidelines’ content should rate their statements by strength according to the quality of evidence provided by well-designed (or not) studies.

Blangis et al² have reviewed 20 clinical guidelines for the early detection and diagnosis of CPA issued in 15 high-income countries. These guidelines were published by academic societies and health agencies between 2010 and 2020 and were dedicated to infants less than 2 years old. The objective of the authors was to systematically evaluate the completeness, clarity, and consistency of these guidelines. Their ultimate goal was to provide orientation for well-designed original clinical studies, systematic reviews, and/or international consensus processes for the diagnostic workup of CPA.

In this innovative approach, each guideline was first classified as a general guidance for early detection and diagnosis of CPA or a narrow scope. Then the detailed contents of definition for sentinel injuries and the recommended diagnostic workup (imaging and laboratory tests) were compared. The contents of the guidelines were compared according to tests gathered in 4 domains: detection of skeletal, head and spine, or thoraco-abdominal injuries and exploration of differential diagnoses.

Among the 20 guidelines retrieved, 4 were issued by radiology societies and they were expected to deal with the imaging tests to perform in case of suspected CPA. The authors found that all other 16 guidelines (those providing general guidance for suspecting CPA) displayed a definition of sentinel injuries. However, the term sentinel injury was only found in 3 of them. While the term sentinel might not be as intuitive in English as it is in French and therefore not broadly used to define lesions other than bruises, it should be remembered that Sheets and al³ have shown that previous sentinel injuries are common in infants with physical abuse and rare in those found not to be abused. If we want to improve detection of occult injuries and ultimately identification of abused infants, we should keep in mind that sentinel injuries are critical to decide which tests to use in the medical evaluation for these infants. The authors also explored the diagnostic and laboratory tests recommended in these 16 guidelines. For the differential diagnosis, recommendations of laboratory tests to explore bone metabolism or bleeding disorders were given, respectively, in 9 and 14 of the 16 guidelines. They also found that the laboratory tests to be performed in case of suspected thoraco-abdominal injury were given in 12 out of 16 guidelines. And finally, all 20 guidelines recommended diagnostic tests to detect skeletal injuries or head and spine injuries and 16 of 20 gave a recommendation for diagnostic tests to perform for thoraco-abdominal injury.

Finally, the authors concluded that main discrepancies occurred with definition of sentinel injuries and performing systematically some imaging tests (bone scintigraphy, spinal magnetic resonance imaging [MRI], cranial ultrasonography [US], chest computed tomography [CT], and abdominal CT and US), whereas the guidelines agreed with recommending radiological skeletal survey, head CT, head MRI, and eye fundus exam but disagreed on whether they should be systematically performed or not. Nevertheless, in my opinion, the most striking finding was that only 3 guidelines have provided details on the quality of evidence and rated the strength of their statements.

In their discussion, the authors suggest that lack of standardized guidelines for the identification and management of CPA may contribute to practice variation, and they give some clues to optimize the early detection and diagnostic workup of CPA. Well-designed original studies and systematic reviews are required for the following clinical questions: bone scintigraphy, spine MRI, cranial US, chest CT, abdominal US and CT, and laboratory tests for abdominal injuries and for differential diagnoses. They also urged to clearly define sentinel injuries and not only their number, location, size, and patterns but also on the specific term sentinel injury as a warning sign. And finally, they pointed out that most of the guidelines for early detection and diagnosis of CPA have not followed international recommendations to elaborate their content. Assuming these propositions, the message is clear: it’s time to go back to work!

As stated by DM Lindberg, “child abuse recognition is the quintessential example of a topic where subspecialty knowledge is needed by a broad range of generalist clinicians,”⁴ and undoubtedly optimal guidelines are key to disseminating this knowledge. Writing medical guidelines is a fascinating task but it is not a simple one. First, you have to elaborate meaningful questions that you will hope to give an evidence-based answer. Second, you have to follow international recommendations for guidelines development processes, and third, you need to work as a team, meaning that you have to discuss almost every word of the guidelines. And finally, you will have to assure the guidelines diffusion to the right targets. After this long journey, you may be satisfied: the work was well done, the objectives were achieved, and your edited guidelines are not only clear and complete, they are optimal. Actually, it is only the beginning since even optimal guidelines, in order to be useful, need also be followed by clinicians. More work for the future!

Published: November 17, 2021. doi:10.1001/jamanetworkopen.2021.30004

Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2021 Boileau P. JAMA Network Open.

Corresponding Author: Pascal Boileau, MD, PhD, University of Versailles St-Quentin-en-Yvelines, Versailles, France (pascal.boileau@ght-yvelinesnord.fr).

Conflict of Interest Disclosures: None reported.