INTRODUCTION

In this study, we review the features of a consultant’s letter most valued by referring physicians. This was prompted by our recent experience with patient and health care administrator engagement in research, leading us to question if we were truly addressing the needs of referring physicians. Letters not targeting the preferences of referring physicians might not provide clear, easily accessible, and actionable information. There is no relevant systematic review.

Referring physicians and consultant specialists work together to provide patient care, and effective communication between physicians is an integral foundation of this relationship. Usually, the communication is conducted through letters exchanged between the physicians. Physicians refer patients with a referral letter and consultant specialists complete assessments and send a written response. There is a broad range of styles, which adds difficulty for readers to extract relevant information with ease and speed. Two recognizable features of the written responses are their structure and their content, for which there is very little guidance.1 As well, we do not know whether referring and consulting share common opinions about valued content and structure.

While the letter style may influence the ease and amount of information extracted by the family physician,2 few physicians report formal training in writing letters.3, 4 Goldman et al.5 first explored the consultation relationship and made ten explicit suggestions to help guide specialist consultations. However, the information on which they were based was descriptive and did not suggest the most valued components of consultation letters. Accordingly, we conducted a systematic review to identify and summarize the key features in a consultant’s letter from a referring physician’s perspective, focusing on its structural and content components.

METHODS

Methodology

A systematic literature synthesis was used to identify and summarize the findings and range of research in the area of consultant letters preferred by consulting physicians. The research question was “What are the components of a consultant’s letter most valued by a referring physician?”

Search Strategy

We searched Embase and MEDLINE (OVID) Medicine (EBM) Reviews and Cochrane Database of Systematic Reviews for English articles with no restriction on initial date to January 6, 2017. Key search terms included the following: communication, correspondence, referral and consultation, letter, inter-professional, physician-physician, doctor-doctor, Family Practice, Family Physicians, General Practitioners, and Consultants. Reference lists of included studies were also searched for relevant citations. Only primary data reports were included in this review. A copy of the updated search strategy is shown in Appendix A.

Study Selection

Two reviewers (VK, AR) independently screened titles and abstracts and removed duplicates. Articles containing letters from specialists to referring physicians regarding outpatient assessments with either an observational or experimental design were included. Studies were excluded if they pertained to communications from referring physicians to consultant specialists, or pertained to allied health professionals, inpatient documents, or opinion reports and editorials (Table 1). Any disagreements were resolved through consensus.

Table 1 Characteristics of Critically Assessed Articles. GP, General Practitioner or Family Doctor; Spec, Specialist
Full size table

Data Extraction

Two reviewers (AR and VK) extracted data using a data extraction form that was developed a priori. A pilot test of the audit tool was conducted to ensure feasibility. Key components of this form included research design, sample size, inclusion/exclusion criteria, data collection methodology and analysis, and main findings. Any disagreements between the two reviewers were resolved by consensus.

Data Synthesis

Following a first review of the assembled literature, we noted they fell simply into three categories: preferences about how information was conveyed, what information was conveyed, and whether referring and consulting physicians shared common opinions about the first two themes. First, we enumerated the frequency with which a theme was addressed, assuming that this reflected interest in the theme. Second, we reviewed the nature of the findings: which specific items were rated positively or negatively?

RESULTS

The identification/screening phase identified 1200 reports. Removal of duplicates and review of titles and abstracts resulted 99 eligible articles. A total of 18 studies finally were included for review and synthesis of information (Fig. 1) Of these, 12 reported themes of letter structure, 8 reported themes of physician preference of letter content, and 4 reported opinions of referring physicians and consultants on letter content.

Figure 1
figure 1

PRISMA flow diagram.

Full size image

Study Characteristics

The characteristics of each article are found in Table 1. The following study designs were represented: observational (n = 15), experimental (n = 2), and systematic (n = 1) review. The earliest report appeared in 1985 and the most recent in 2013, with 14/18 published in 1993–2010. All were from Commonwealth countries (the UK, eight; Australia, six; Canada, three; Singapore, one). The consultant specialties included oncology five, various internal medicine three, pediatrics two, otorhinolaryngology two, dermatology one, psychiatry one, and not specified four. Seventeen articles utilized surveys as the main method of data acquisition. Eight articles included use of an audit tool to evaluate consult letters. One article1 used a Delphi consensus process and did not report quantitative data.

Letter Structures

Twelve articles reported on letter structures.1, 2, 6, 7, 10, 13,14,15,16,–17, 19, 20 Three major elements were identified: structured letters (n = 10), itemized lists (n = 4), and length of reports (n = 4). All ten articles1, 2, 6, 7, 13,14,15,–16, 19, 20 covering whether reports should be structured noted preference for this format. Table 2 reports the proportion of respondents who preferred a structured format to all other formats combined; the mean of the means of individual reports was 87 ± 9% and the mean of all respondents was 91.4%. Not surprisingly, four papers1, 10, 13, 14 reported that referring physicians also expressed a preference for itemized lists in the consultant’s letter. These included diagnoses, test results, co-morbidities, medications, and management plans. The amount of information contained in the letter was a commonly recurring element, with three articles discussing it.9, 15,16,–17 All four preferred brevity although none addressed this quantitatively. Two mentioned lengths of less than two typewritten pages,17 or less than 350 words.16

Table 2 Proportion of Family Doctors Preferring a Structured Letter to All Other Options Combined. The Mean of All Included Subjects Was 91.4%
Full size table

Letter Contents

Only 8 of 18 articles6, 8, 9, 11, 12, 17, 18, 21 provided information on preferences of referring physicians about content (Tables 1 and 3). Three themes were identified: Formulation, Treatment, and Data. The most common subthemes were in Formulation: diagnosis, management plan, and prognosis; in Treatment: treatment and treatment options, side effects, and what the patient was told; and in Data: test results, physical examination, history, current medications, and patient’s wishes. We then scored the ranking of the themes in each article in the top, middle, or bottom tertile (Table 3).

Table 3 Preferences of Referring Physicians for Contents of Letters from Consultants. The Preferences in Each of Eight Relevant Reports Were Ranked, Then Grouped into Tertiles. For Example, Diagnosis Was Included in All Eight Reports and Was in the Top Tertile in Seven and the Second Tertile in the Remaining Paper
Full size table

The three items in Formulation were reported 16/21 times in the top tertile; the three items in Treatment were reported 4/15 times in the top tertile; and the five items in Data were reported 1/23 times in the top tertile. Conversely, the three items in Formulation were reported 2/21 times in the bottom tertile; the three items in Treatment were reported 3/15 times in the bottom tertile; and the five items in Data were reported 13/23 times in the bottom tertile. Therefore, referring physicians appear to value formulation and insight over treatment options, and the latter over conventional data base elements.

Opinions of Referring Physicians and Consultants

Seven reports touched on whether referring physicians and consultants shared preferences and values in letters by consultants. Of these, three were unhelpful. Berendsen et al.22 focused on inter-professional opinions of referral and consulting style, Keely et al.9 had only eight rankers whom four were family doctors, and Tattersall et al.18 presented no relevant data. Of the remaining four studies, two reported no differences in opinions by referring doctors and specialists. Newton et al.12 found no significant difference between specialists and general practitioners when both groups were surveyed regarding items in a consultant letter that they deemed to be always or usually important. Westerman et al.23 used only eight rankers (four specialists, four family doctors), who agreed on the values of all aspects in consultants’ letters.

Three reports did find differences of opinions between family doctors and specialists. McConnell et al.11 noted that family doctors and surgeons identified similar information gaps in letters from oncology specialists but differed in the preferred amount of information. Referring family doctors wanted more detail in all aspects of the letter, while referring surgeons preferred a crisper version. Superfluous information was felt to be a problem more commonly by surgeons than by family doctors. The referring physicians wanted details on treatment, management, and prognosis, and these details were frequently absent. In contrast, background information was liberally provided by oncologists and valued less by referring physicians. Although there were numerous comparisons with apparent differences, their individual statistical significances were not reported. Babington et al.6 compared the opinions of referring family doctors and specialists on the value of medical oncology consultation letters. Compared to consultant specialists, referring family doctors were significantly more likely to prefer information on management plan, test results, treatment side effects, current medications, and discussions with the patient.

DISCUSSION

Main Findings

This is the first broad review on the topic of components of a consultant letter preferred by referring physicians. Referring physicians overwhelmingly prefer a brief, structured report that features value-added insight: diagnosis, prognosis, and management plan. Of less interest are the primarily data fields such as history, physical examination, and current medications.

Structure of Consultation Letter

Referring physicians favor succinct, structured letters10, 14 containing itemized list of identified problems and management suggestions (Table 2). Reasons for this preference may include ease of access to information with subheadings, concise content, and clarity in presentation.15 For example, comprehension is improved in structured letters with problem lists compared to unstructured letters.2 This includes computerized structured letters that contained subheadings and concise information. Structured letters take less time to read and increase reader recall compared to unstructured letters.2

Content of Consultation Letter

Referring physicians consistently favored the value added by consultation: diagnosis, prognosis, and management plans feature highly in preference lists (Table 3). Of medium importance were the details of the treatment plan, and of much less importance were conventional data features: history, physical, current medications, and patient preferences. The reports favored the synthesis, insight, and formulation of consultants over including the often voluminous data that underpins the synthesis.

Differing Perspectives of Referring Physicians and Consultants

It may be that specialists and referring physicians do not completely share preferences and values in consultation letters to family doctors. The weak and conflicting data preclude a stronger conclusion. If true, they may disagree because the letter serves a different function for each party.9 Consultants may use the letters as a detailed historical recollection of the patient encounter24 and therefore, as a detailed database for future care. Although much of this information is already known to the referring physician, some more specialized and detailed information pertinent to the referring physician’s request may be unknown to them. Expectations of a letter can also change based on audience; for example, surgeons and GPs desired similar information, with different amounts of detail.25

Main Implications

Referring physicians appear to place high value on letters being succinct, clear, and organized, and possibly with a standardized template. They value a management and follow-up plan more than a detailed recitation of the history and investigations. In short, they prefer a clearly visible opinion and plan. This suggests that the commonly used narrative review interspersed with lists of co-morbidities and medications may not be as preferred a mode of communication as is commonly assumed. Consultants’ letters may simply be following human nature to review, consider, formulate, and plan. Implementing family physician preferences may improve the perceived value of letters. Of utmost importance is conveying clear management plans, what the patient has been told, and prognosis.

Limitations

The heterogeneity of the reviewed studies precluded us from utilizing previously developed critical appraisal tools. Additionally, due to the marked heterogeneity of study design and reporting, we were restricted to a descriptive analysis. Although a thorough search of the literature was done, it is possible that studies may have not been included due to being published not in English. All the reports are from Commonwealth countries and whether the conclusions are relevant to the American style of health care delivery, funding, physician reimbursement, and liability is unknown. The studies we chose to review were from specialists to referring physicians in an outpatient setting; therefore, our findings may not be applicable to an inpatient environment. It is also worth noting that a significant amount of the literature comes from oncology backgrounds. Indeed, most of the studies were based on single consultants, single specialties, or single sites. The dates of the studies span the last 30 years, and only one includes mention of computerized information. Therefore, we cannot comment on the evolving utility of electronic medical records, voice recognition software, or social media. Interestingly, however, the same themes recur over 30 years, suggesting that their transcendent importance awaits a solution. This only includes one aspect of consultation, whereas face to face, co-management and other aspects are important in this relationship, but were beyond the scope of the current study.

Conclusions and Future Directions

Future studies might first pilot investigations of optimum templates and structures, then perform subsequent randomized studies of knowledge retention and change in practice. Solutions will differ depending on broad categories, such as surgical versus internal medicine opinions, first consultations versus follow-up letters, and initial opinions versus transfer of care. The solutions to these problems are probably best addressed with an integrated knowledge translation approach, in which the target knowledge users (referring physicians) participate in framing the questions and methodological approach. Finally, the different needs of referring physicians and consultant thinking process may require post hoc processing of dictations into cued templates. This report may result in observations on implementing Competency 3 (communicate effectively with physicians, other health professionals, and health-related agencies) in the Domain of Competence: Interpersonal and Communication of the Accreditation Council of Graduate Medical Education Skills New Accreditation System {Benson, 2014, Acad Pediatr, 14, S55–65}.