BMJ Qual Saf 20:692-697 doi:10.1136/bmjqs.2010.045781
  • Original research

Development and validation of a tool to improve paediatric referral/consultation communication

  1. Richard C Wasserman4
  1. 1Department of Pediatrics, University of Colorado School of Medicine, Aurora, Colorado, USA
  2. 2Department of Medicine, Meyers Primary Care Institute, University of Massachusetts Medical School, Worcester, Massachusetts, USA
  3. 3Meyers Primary Care Institute, University of Massachusetts Medical School, Worcester, Massachusetts, USA
  4. 4Department of Pediatrics, University of Vermont College of Medicine, Burlington, Vermont, USA
  1. Correspondence to Dr Christopher Stille, University of Colorado/The Children's Hospital, 13123 East 16th Avenue, B032, Aurora, CO 80045, USA; stille.christopher{at}
  • Accepted 25 January 2011
  • Published Online First 21 February 2011


Objective To develop a template to promote brief but high-quality communication between paediatric primary care clinicians and consulting specialists.

Methods Through an iterative process with academic and community-based paediatric primary care providers and specialists, the authors identified what content elements would be of value when communicating around referrals. The authors then developed a one-page template to encourage both primary care and specialty clinicians to include these elements when communicating about referrals. Trained clinician reviewers examined a sample of 206 referrals from community primary care providers (PCPs) to specialists in five paediatric specialties at an academic medical centre, coding communication content and rating the overall value of the referral communication. The relationship between the value ratings and each content element was examined to determine which content elements contributed to perceived value.

Results Almost all content elements were associated with increased value as rated by clinician reviewers. The most valuable communications from PCP to specialist contained specific questions for the specialist and/or physical exam features, and the most valuable from specialist to PCP contained brief education for the PCP about the condition; all three elements were found in a minority of communications reviewed.

Conclusions A limited set of communication elements is suitable for a brief communication template in communication from paediatric PCPs to specialists. The use of such a template may add value to interphysician communication.


In the ideal ‘medical home’ model of care,1 promoted in the USA as the model of high-quality primary care for the future,2 the primary care practice is typically the ‘hub,’ where care is coordinated and where a working partnership is developed between primary care providers, specialists, community resources and families. Within this partnership, communication between providers in different settings is essential to ensure that needs are met while avoiding duplicative care. Effective ‘handoffs’ between care settings are also critical in ensuring safe care and avoiding medical errors.3

Communication between paediatric primary care providers (PCPs) and specialists is frequently a problem, with delayed or absent communication and insufficient discussion of shared responsibility two important deficits.4–6 Part of the problem is the lack of a system to promote routine, efficient communication for most primary care practices.4 7 Lack of reimbursement for activities involved in primary care-specialist coordination of care further contributes to inconsistent communication.8

Previous studies found that communication from PCPs to paediatric medical specialists for newly referred patients was relatively infrequent.5 6 Further, specialists frequently reported difficulty providing optimal care when communication from the PCP was absent.6 In response, to develop the intervention for the current study, we worked with physicians to design a brief template that could be used by both PCPs and specialists in their busy office environment to improve communication. The goal was to maximise the presence and quality of documented, bidirectional communication between paediatric PCPs and paediatric specialists about new referrals. To minimise extra work while maximising value, we worked with physicians participating in the study to develop a short list of the most important content elements for communication from PCPs to specialists as well as from specialists back to PCPs.

In an effort to validate the communication template as developed by physicians, we examined the relationship between communication content elements contained in the template (such as medical history or current medications) and the communication's value as rated by trained clinician reviewers, in a sample of actual patient referrals from community paediatricians to specialists at an academic medical centre.


Setting and intervention design

This study was conducted in 30 paediatric primary care practices and five paediatric specialty practices in both academic and community settings in Central Massachusetts, as part of a larger observational study of primary care–specialty communication. Details of the larger study, and description of physicians and their practices, are described elsewhere.6

Participating physicians were invited to one of three 1 h meetings to identify elements of communication that were important for patient care, yet brief enough to communicate in the context of daily practice. Physicians were asked what the content of ideal, brief communication should be, and their preferences were noted. These discussions, together with a review of previous literature, 5 7 9 10 were used to generate a candidate list of communication content elements, attempting to capture those elements described as most important while avoiding duplication and unnecessary detail. The list was circulated once again among the physicians, to make sure it reflected their views, and was approved by consensus with minor revisions.

The final list consisted of six key ‘referral’ (PCP to specialist) and 12 key ‘consultation’ (specialist to PCP) content elements (see box 1).

Box 1

Important communication elements

From primary care provider to specialist (‘referral’):

  • Reason for referral

  • Specific questions for the specialist

  • History of the condition

  • Medications

  • Physical exam features

  • Any diagnostic testing done

From specialist to primary care provider (‘consultation’):

  • History of the condition

  • Physical exam features

  • Mention of any diagnostic testing (or no testing)

  • Diagnosis

  • Reasoning behind the diagnosis

  • Description of immediate management

  • Description of longer-term plan

  • Reasoning behind the plan

  • Medications

  • Discussion of sharing of care with the primary care provider

  • Education about the condition

  • Communication with any other specialists involved in care

While desired content determined the final number of items on the list, many physicians commented that any feasible form would have to be extremely brief. Therefore, all elements were incorporated into a one-page template (figure 1), with essential PCP to specialist (‘referral’) details on the top half and critical specialist to PCP (‘consultation’) details on the bottom half.

Figure 1

Referral/consultation template

Patient eligibility and enrolment

Eligible subjects were children enrolled in participating primary care practices who were seeing a participating paediatric specialist for the first time. These children were identified by review of upcoming schedules at the specialty practices 2 weeks before the visit. Parents of eligible children were sent a letter introducing the study and informing them that a researcher would be speaking with them at their upcoming specialty visit. This letter provided a phone number to call if the parent did not wish to be included in the study. Interested parents provided informed consent at the first visit, then answered a brief demographic questionnaire.

Data collection

Deidentified copies of the child's PCP and specialist charts, obtained 4 weeks after the initial specialty visit, were reviewed independently by two trained, actively practising paediatric nurse practitioners (one specialty, one primary care) who were not involved in the intervention. Training lasted 2 h on average and consisted of a review of detailed descriptions of each content element, with examples of what text did and did not qualify, followed by grading of 10 pilot charts with feedback from the PI. Reviewers looked for the presence of any documented communication from the PCP in specialty charts, and any documented communication from the specialist in PCP charts. Each reviewer graded the communication from each chart for how valuable overall they felt it would be if they were the receiving clinician, on a 5-point Likert scale (1=not valuable; 5=very valuable); scores were averaged to determine the final result. Only charts with communication present were included in the average score. Reviewers also noted the presence or absence of key content elements in each chart for which communication was documented; disagreements between reviewers were decided by the first author to determine the final result.


To determine the relationship between the presence of different content elements and value of communication, the average value of all charts containing each element was compared with the average value of those that did not using t tests. For example, the average value of the 25 PCP to specialist communications that contained a specific question was compared with the average value of the 64 PCP to specialist communications that did not. Statistical significance was determined using a threshold of α=0.05. Inter-rater agreement on the presence or absence of content elements was determined using the Cohen κ. Data were analysed using SPSS, version 16.0.

The study was approved by the institutional review boards of the University of Massachusetts and the Fallon Clinic.



Patients were cared for by 62 different PCPs in 30 practices, and 15 specialists in five specialties: gastroenterology, pulmonology, cardiology, endocrinology and neurology. Demographic characteristics of the patients are presented in table 1.

Table 1

Patient characteristics

Documented communication, and its value

Charts from 206 patients were reviewed; 89 had communication from PCP to specialist, and 182 had communication from specialist back to PCP. The value of all communication from PCP to specialist, averaged between the two reviewers, was 3.0±1.1 on a five-point scale. The average value of all communication from specialist to PCP was 3.9±0.7.

Content of communication

On average, there were 3.3±1.5 out of a possible six ‘referral’ (PCP-to-specialist) elements in each chart with documented PCP-to-specialist communication. Proportions of charts with communication in each group that contained each element are listed in table 2. Elements included most frequently were diagnostic testing, reason for referral and history of the problem (each in over 70% of charts with documented communication); a specific question for the specialist was least frequently included (28%). The inter-rater reliability for each element was moderate (κ≥0.4) or better, with the exception of the reason for referral (κ=0.18).

Table 2

Comparative value of referral communication (primary care provider to specialist, n=89) containing different content elements

On average, there were 8.1±1.7 out of a possible 12 ‘consultation’ (specialist to PCP) elements in each chart with documented specialist-to-PCP communication. Proportions of charts with communication in each group that contained each element are listed in table 3. Elements included most frequently were history, physical exam findings and an initial management plan (each in at least 97% of charts with documented communication); education for the PCP, mention of sharing of care with the PCP, and mention of sending communication to other providers involved in care were least frequently included (all under 30%). The inter-rater reliability for each element was variable, with κ values between 0.18 and 0.89; all disagreements were resolved by the first author.

Table 3

Comparative value of consultation communication (specialist to primary care provider (PCP), n=182) containing different content elements

Relationship between content and value

There was a strong positive relationship between the presence of most content elements and the value of the communication. In referral communication (from PCP to specialist), each content element except mention of diagnostic testing was associated with increased value, with the highest value observed in charts with specific questions for the specialist and physical exam features present (table 2).

The inter-rater consistency for value on the 1–5 scale was good, with raters agreeing within one point 60% of the time and within two points 97% of the time. In consultation communication (from specialist to PCP), nine of 12 content elements were associated with increased value, with the highest value observed in charts with education for the PCP about the patient's condition (table 3). Raters agreed on the value of consultation communication within one point 94% of the time.


As part of a project to improve the quality of communication between referring and consulting physicians in ways that are feasible to undertake in the context of daily practice, we sought to develop, evaluate and validate what clinicians determined to be value-added elements of communication. Unlike previous studies, the list was designed a priori, in collaboration with participating physicians, to improve its usability in practice and its face validity for practice-based research; additionally, the list was formulated for communication from both PCP to specialist and specialist back to PCP. Presence or absence of content elements in communication was based on objective evaluation by experienced paediatric primary care and specialty clinicians not involved in care of the patients in the study.

The list of communication content elements chosen for this study appears to be appropriate. When the content of communication was compared with its overall value as rated by the clinician reviewers, most of the six ‘referral’ (PCP to specialist) and 12 ‘consultation’ (specialist to PCP) content elements that were chosen in collaboration with participating physicians contributed to perceived value and constituted a reasonably parsimonious list. Those that were not associated with increased value (previous testing for referral communication, and medications for both types of communication) are important to minimise duplication of care and for reasons of patient safety. All five content elements found to be associated with quality of communication, paediatrician satisfaction and/or benefit to patient care in a previous study5 (history, diagnosis, management plans, follow-up plans and discussion of comanagement) were also associated with increased value in our study. Referral communications with the greatest value included those with specific questions for the specialist, judged to be essential in a number of previous studies5 but present in only 28% of communication in our sample,9 10 and physical exam features. These elements likely serve to orientate and prepare the specialist, and have the potential to make the consultation more efficient. Consultation communications with the greatest value included those with education for the PCP, present in only 27% of communications. This is important when considering the Medical Home model, as it enables the PCP to participate in comanagement and discuss expectations. Plans for comanagement were the top predictor of PCP satisfaction in a previous large study.5 These results suggest that the use of a template, if it served to prompt providers to include items such as specific questions for the specialist or education for the PCP, could increase the value of communication substantially. However, the larger problem of lack of communication in the first place, as evidenced by fewer than half of specialist charts containing communication from the PCP, remains to be solved. If a fax-back form, promoted by experts in Medical Home quality improvement,11 were to be used, this may help; alternatively, a tool embedded within an electronic medical record has been shown to be promising in a single-site study.12


It is important to note that optimising the content of information transferred is only one part of effective communication; providers still need to communicate in a timely manner that addresses the needs of the patient/family and care team. However, striving for a brief, useful list of information may help to streamline the communication process to make it more effective. Methodologically, there was low inter-rater reliability in judgements for a few content elements. Resolution of disagreements by the principal investigator likely mitigated some of this effect, and our technique represents an improvement over earlier studies that relied on physician self-report.


When this study is combined with other literature, it appears that the essential content elements for paediatric referral communication are now well established. Additional strategies are needed to integrate communication into routine referral procedures; this will likely require larger practice process changes that may incorporate a communication template.


We thank the physicians and staff at the participating practices for support in completing this study, J Grimes and N Hagberg for assistance in chart reviews, and A Giese and C Gagne for expert technical assistance.


  • Funding This work was supported in part by a grant from the Robert Wood Johnson Foundation Generalist Physician Faculty Scholars Program (#42206).

  • Competing interests None.

  • Ethics approval Ethics approval was provided by the University of Massachusetts Medical School and the Fallon Clinic.

  • Provenance and peer review Not commissioned; externally peer reviewed.


Free sample

This recent issue is free to all users to allow everyone the opportunity to see the full scope and typical content of BMJ Quality & Safety.
View free sample issue >>

Email alerts

Don't forget to sign up for content alerts so you keep up to date with all the articles as they are published.