Taking the Pain out of Somatization: Development and Implementation of a Hospital-Based Clinical Practice Guideline to Address Pediatric Somatic Symptom and Related Disorders

This is a single-center quality improvement and feasibility study developed and implemented at a 230-bed academic pediatric hospital with active child life, pediatric psychology and psychiatry consultation, palliative care, and adolescent medicine services. There is access to an on-site psychiatric ED but no access to a mental health consultation service in the pediatric ED. This study began in June 2014, was completed in September 2015, and was granted a quality improvement exemption from the hospital’s institutional review board.

The multidisciplinary group developed from the Conversion Disorder Journal Club, which met every other month from May 2013 to May 2014. Group membership was broadened into a larger workgroup given the increasing recognition of the needs and challenges related to the care of SSRDs. The multidisciplinary group included representatives from psychology, neurology, hospitalist medicine, psychiatry, primary care, physical medicine and rehabilitation, EM, physical therapy, occupational therapy, speech therapy, social work, child life, hospital administration, school services, and nursing as well as parents. The purpose of the workgroup was to better understand current SSRD practices and to develop a standardized evidence- and consensus-based CPG and protocol with resources for the inpatient and ED settings. This group formally started meeting in June 2014 and met 20 times, approximately monthly, between June 2014 and December 2015. The protocol was launched on October 15, 2015. Representatives from pediatric hospital medicine, psychology, psychiatry, neurology, and EM continue to meet quarterly to ensure fidelity of the interventions and make protocol adjustments on the basis of iterative plan-do-study-act cycles.¹⁶ 

Two methods were used to evaluate the current state and inform protocol development: retrospective chart review to characterize current practice and lean methodology with value stream mapping (VSM) to better identify problem areas within clinical care, workflow, and systems of care.¹⁷ 

A retrospective chart review was performed on a convenience sample of patients with a suspected or confirmed SSRD admitted between May 2012 and October 2014. No consistent methodology previously existed to identify patients with SSRDs because of limited recognition and diagnostic practice. Patients were identified through previous clinical interaction by members of the workgroup providing names of patients with any suspected diagnosis falling under the SSRDs diagnostic category. Patients were excluded from analysis if they were only seen as outpatients, if they were discharged from the ED, or if there were insufficient data to analyze. If multiple admissions occurred, the most recent admission was selected for review.

Electronic medical records were reviewed by a pediatric resident, a pediatric hospitalist, and an undergraduate student to obtain the following: age, sex, race, ethnicity, chief complaint, admission service, route of admission, length of stay (LOS), imaging studies, procedures, service(s) consulted and consultation date, number of discharge diagnoses, whether the discharge diagnosis included an SSRD, psychiatric comorbidities, recommended follow-up, and discharge medications.

A lean coach assisted with use of lean methodology, an approach to workflow and resource management aimed at reducing waste and improving efficiency and quality.¹⁷  VSM is a lean management method for analyzing the current state to identify areas of inefficiency in a system that may serve as points of intervention in a future state.¹⁷  Multidisciplinary group members used VSM to define the flow of SSRD care for a given patient from the ED to discharge, noting multiple areas of inefficiency and disorganization (Fig 1). This included points of communication with the patient and family and opportunities for multidisciplinary collaboration. Common barriers to flow and problematic issues from each point of care were discussed. Members identified areas of waste in the process and ranked problematic areas, considering feasibility of change, severity of need, and importance of the problem. This process aided in prioritization of problem areas as well as a potential future solution. In Fig 1, items noted within the thunderclouds highlight these challenges, whereas the tornadoes identify areas of chaotic workflow and potential waste.

Three workgroups were created: (1) family education, (2) provider education, and (3) protocol and CPG development. Parent advisers provided feedback on the protocol and all materials given to families. The larger group reviewed and approved clinical and educational resources, including the protocol and CPG for the health system.

Sixty patients were initially identified for chart review. Twenty-four were excluded from analysis (15 discharged from the ED, 6 with insufficient medical records, and 3 outpatients). The 36 patients included in the final analysis ranged in age from 8 to 17 years old (mean = 13.5, SD = 2.6). The majority were girls (61%), white (78%), and non-Hispanic (97%).

The most common chief complaint was categorized as neurologic (50%), followed by pain (22%), gastrointestinal (17%), and cardiopulmonary (11%). The majority of admissions were from the ED (61%), whereas 22% were scheduled admissions from an institutionally affiliated outpatient provider, 11% were from the post–anesthesia care unit, and 3% were direct admissions from a subspecialty clinic. Most patients were admitted to the pediatric hospitalist service (42%), followed by the neurology (39%) and gastroenterology (6%) services. The remaining 5 patients were each admitted to 1 of the following services: physical medicine and rehabilitation, surgery, rheumatology, neurosurgery, and critical care medicine. The mean LOS was 5.44 days (SD = 6.3, range 1–33 days). The mean number of imaging studies obtained per patient was 1.0 (SD = 1.1). The mean number of nonimaging diagnostic procedures performed per patient was 1.0 (SD = 1.0).

Eighty-six percent of patients received a consultation from at least 1 of 17 subspecialty services (Table 1). The mean number of consultations per patient was 2.3 (SD = 2.0). Of those who received consultations, 23% received a consultation on hospital day 1. Seventy-seven percent of patients received a psychiatry and/or psychology consultation, with only 1 of these consultations occurring on hospital day 1.

At discharge, patients had a mean number of 5.8 (SD = 5.2) medical and/or psychiatric diagnoses, with only 33% having any formal SSRD diagnosis. Patients were prescribed an average of 2.1 (SD = 2.6) new medications; the mean number of discharge medications was 6.3 (SD = 4.4). The mean number of comorbid psychiatric diagnoses was 0.9 (SD = 1.2). Forty-seven percent of patients had psychiatric comorbidities, including depressive disorders (59%), anxiety disorders (59%), attention-deficit/hyperactivity disorder (35%), autism spectrum disorder (18%), posttraumatic stress disorder (12%), oppositional defiant disorder (6%), bipolar disorders (6%), and personality disorders (6%).

The mean number of follow-up appointments recommended per patient was 2.2 (SD = 1.7), with 4 patients receiving no follow-up recommendation (11%). Appointments were recommended with 20 different subspecialists. Only 53% had a recommendation for primary care physician follow-up. Mental health services were recommended for less than half of patients, with 47% referred for psychotherapy and 28% for psychiatry. Twenty-five percent of patients received no mental health follow-up recommendations. Physical medicine services (ie, physical therapy, occupational therapy) were recommended for 25% of patients. One patient (3%) was transferred to inpatient psychiatry, whereas all other patients were discharged from the hospital.

On the basis of chart review and VSM findings (Table 2), a protocol was developed to guide the evaluation and management of a patient with a suspected SSRD (Figs 2 and 3). This protocol integrates physical health care and mental health care throughout the evaluation process, with greater awareness of a possible SSRD diagnosis starting in the ED. The protocol encourages a thorough workup along the biopsychosocial continuum, early mental health involvement, and caution with invasive, low-yield interventions and diagnostic testing. It includes provisions for verbal and written information for families, including standardized educational handouts (available on request). At the time of the development and implementation of the protocol, mental health services were not directly available to the medical ED setting, but ED providers and social workers were encouraged to call on-call mental health providers if questions existed. Discussions are ongoing regarding establishment of a mental health service in the medical ED, with a proposed plan to pursue this in the spring of 2020.

A clear discharge plan and resources, including mental health referrals, are outlined, with guidance on communication between providers. Informed by the consultation liaison psychiatry and psychology services, as well as by social workers in the department of psychiatry, a referral list is available to the ED and inpatient services to provide mental health referral to patients for SSRD care. Social workers will see patients in the ED before discharge to review mental health and other referral resources and provide added education to patients and families. In complex cases, the social work in psychiatry service will provide additional support and mental health referral. Because social workers and providers cannot enroll patients in mental health care directly, families are provided the referrals and key contact information but are then expected to make the mental health appointments themselves. In some circumstances, ensuring that mental health follow-up is in place is a criterion for discharge in either the ED or the inpatient setting.

The CPG was developed as a brief narrative document detailing the principles and practices employed within the protocol (available on request). The protocol, CPG, and all associated resources are readily available on the hospital intranet Web site, and a protocol flow sheet is posted in each resident physician workroom.

During the first year of implementation, several changes were made to the protocol-based information gathered from plan-do-study-act cycles and provider feedback.¹⁶  Social work consultation was withdrawn as a standard order because of clearer options for discharge referrals and overlap of role with psychology and psychiatry services. School services were no longer automatically requested because youth were not staying in the hospital long enough to need help with school reentry or bedside instruction. More specific guidelines to the protocol were added to guide decision-making when consulting subspecialty and mental health services.

Protocol implementation included formal education of pediatric resident physicians, pediatric faculty, nursing, and staff in the ED and inpatient settings. An e-mail correspondence was sent to pediatric faculty and staff outlining the protocol, CPG, and associated resources. A psychologist, psychiatrist, neurologist, and hospitalist from the multidisciplinary workgroup provide pediatric residents with annual education, including didactic instruction, case-based role-play, communication strategies, and opportunity for discussion. Feedback and questions are relayed to the multidisciplinary workgroup leadership and are reviewed and addressed just in time or during routine multidisciplinary workgroup meetings.

Although SSRDs are common in inpatient pediatric settings, there has been limited attention to the development of standardized protocols to evaluate and treat these disorders.^1,2  Recommendations for communication and intervention have been suggested without clear guidelines for translation into practice.^4,7,11  This is best exemplified in the recently published AACAP PaCC SSRD pathway.¹⁵  Because these broad recommendations do not address the unique issues present in any 1 health care system, it is imperative to develop and operationalize system-specific interventions for youth admitted with SSRDs. With this study, we present the first institutionally based effort toward standardization of the evaluation and care of youth with suspected SSRDs from the ED to discharge using retrospective chart review, lean methodology, and the expertise of a multidisciplinary team of stakeholders.

Key findings during the review of the current state, outlined in Table 2, were used to depict a more ideal workflow to standardize the evaluation and management of inpatient SSRD care. This study highlights the use of lean methodology to develop targeted clinical interventions, including protocol and CPG development. The lean process and VSM identified areas of waste and systems challenges that might have otherwise been missed if an overly solution-focused approach was pursued. This process encouraged a multidisciplinary approach and the crucial involvement of stakeholders. Particularly unique was the inclusion of parent representatives in our protocol development because they provided a needed perspective on the family experience. Our consensus development process matched the multifactorial nature of SSRDs and the need for a multidisciplinary approach to SSRD care.¹ 

The standardized approach of the SSRD protocol allowed for the creation of data points to assess the impact of this process. The main strength of our protocol is its early initiation in the patient encounter and an expectation for joint medical and mental health care. As a result, the SSRD diagnosis is on the differential early, with a close alignment of expectations for care in communication with families; the protocol and associated resources prompt the clinical team to communicate a diagnostic conceptualization early and document diagnoses in the electronic medical record, both critical steps in SSRD care.¹¹  Expectations for the admission are discussed early and aligned across the spectrum of care to decrease the risk of fractionated care. Mental health providers are consulted after admission, with the goal of providing a full diagnostic assessment, reducing unnecessary invasive cares, and providing a collaborative approach to care to minimize families’ perceptions that they are being “handed off to psych.” Early mental health involvement is important to avoid low-yield tests by focusing on biopsychosocial factors impacting symptom presentations and interventions and to move forward with a plan that effectively treats the SSRD.¹⁴ 

Although we created a process that was unique to our institution’s needs, it came with significant challenges. It is difficult to shift from a culture of conceptualizing somatization as solely psychological and outside the realm of physical medicine to one in which somatization is conceptualized as a disease process that requires integrated management at all levels of care. Inclusion of SSRDs on the differential and the integration of mental health early in the process, simultaneous with ongoing medical workup, was a dramatic shift in practice. These challenges highlight the importance of collaboration across disciplines, with close observation, multidisciplinary care, and ongoing evaluation with the evolution of symptoms. This also includes a measured approach with close communication between inpatient and outpatient providers to ensure consistent communication and longitudinal monitoring and management of the multifactorial nature of SSRDs.¹¹  Multidisciplinary buy-in requires significant relationship building, collaborative communication, and a strong knowledge of existing evidence and current practice. In a large academic health care system, it can be cumbersome to have consistent messaging across providers given multiple levels of providers and trainees. This makes intervention fidelity challenging, which is mitigated by the availability of standardized resources, including the protocol and CPG.

There are several factors that limit the generalizability of this study. The protocol was developed at a single site and may not be applicable to other pediatric settings. However, the principles and general approach should be generalizable in most hospital settings that have access to mental health consultation services. In addition, the patients selected for chart review, which informed our understanding of current practice, were selected by provider recall. Selection bias may have led to inclusion of patients with more disease severity and may not be representative of our general inpatient population with SSRDs. Furthermore, the convenience sample size was small, making interpretation challenging outside of informing CPG and protocol development. However, the characteristics of our patient group do not appear to deviate from those in other published reports of youth admitted with SSRDs.⁴  In this study, we did not use a formalized consensus gathering process, such as the Delphi procedure.^19,20  Although we used lean methodology to identify and address waste in our system, the consensus gathering process was not blinded and could have introduced bias from more-vocal members of the multidisciplinary group.

Future directions for the study of the management of SSRDs in the pediatric hospital setting should be focused on assessment of clinical outcomes.^1,11  There is little literature on multidisciplinary service delivery and whether such programs lead to positive outcomes for this population.¹¹  Given the resources needed to develop and implement this SSRD pilot inpatient intervention, the issue of cost should be addressed. With the considerable cost incurred by nonstandardized SSRD care and the impact of a coordinated approach with early mental health involvement and reduced LOS, we suspect that cost and use should be significantly reduced by the standardization of care.^6,8,14,21  Demonstration of cost reduction is crucial for securing future support of hospital administration in a challenging health care environment.²¹  Our program is currently evaluating our clinical outcomes to be published at a later date. Finally, future study of youth with SSRDs should include the perspective of patients, parents, and medical staff and how a standardized approach, such as the one described in this study, may impact attitudes, perceptions, and future interactions within the health system.

In this study, we describe the development of a multidisciplinary workgroup at a large academic pediatric hospital to standardize the evaluation and care of youth with suspected or confirmed SSRDs in the inpatient pediatric care setting. Our local efforts are in line with the recently published AACAP PaCC SSRD pathway.¹⁵  This single-site study provides a promising approach to standardizing SSRD care and may provide a better means of assessing the quality of care and outcomes for youth admitted with SSRDs.

We thank all those involved on the SSRDs workgroup for their input and dedicated effort to improving care for this population: James Azim, JD, MD, Jeanne Seyfried, MD, Joe Hornyak, MD, Liza Green, MD, Rita Ayangar, MD, Samual Neher, MD, Louis Dang, MD, PhD, Laura Taylor, MD, Tracy Klein, Chris Tapley, Beth Riske-Rogers, Carol Fanning, Nora Rosenblum, Jeremy Gluskin, Kim Moosekian, Kelly Parent, Kori Jones, Karen Adkins-Bley, Corrie Pennington-Block, David Cruz, Stephen Fischer, Nina Steinberg, and Theresa Gattari. We thank our Patient-Family Advisory Council for generously providing funding to support purchases of the educational book for families.