Referral letter: what it must tell the specialist

A 61-year-old man arrives at the specialty outpatient clinic with a folded sheet of paper in his hand. On it is written only the name of an organ and the word assessment. The specialist does not know which symptom prompted the request, which tests have already been done, what the family doctor suspects, or what is expected from the consultation. The investigation starts from scratch. Weeks later, the family doctor still does not know whether the patient was seen.

The referral is one of the most fragile moments in care. It is there that responsibility passes, fully or partially, from one professional to another, and it is there that information is most lost. This text discusses when to refer, what the letter needs to contain, why the specialist's response matters as much as the request, what research shows about interventions to improve this communication, and how to follow the referral through to the end.

Key points

• The threshold for referring varies widely among doctors, which leads to both underuse and overuse of the specialist.
• Many letters arrive without enough information, and many doctors do not know whether the patient was ever seen.
• A good letter states the reason, the question, what has already been investigated, the treatment in use, and the degree of urgency.
• The specialist's response is part of the referral; without it, continuity is lost.
• Checklists, structured templates, and peer feedback improve the quality of letters.
• Following the referral through to the response is the task of the one who refers.

Summary

• Why referrals fail
• The filter role of primary care
• Deciding whether and when to refer
• What the letter needs to say
• Templates and checklists
• The specialist's response
• Following the referral
• When the relationship between levels fails
• Frequently asked questions
• Final checklist

Why referrals fail

An analysis of referrals to specialists used an athletics image to describe the problem: the baton is dropped at the handoff (Mehrotra, 2011). The authors gathered data showing failures at every stage of the process, from the decision to refer to the return of information.

The first failure is in the decision itself. The referral threshold varies so much that some primary care doctors refer more than five times as often as others (Mehrotra, 2011). This variation produces two problems at once: people who would need the specialist do not reach them, and people who would not need them take up slots (Mehrotra, 2011).

The second failure is in the information. Many referrals arrive with no clinical information at all and, when it exists, it is often insufficient (Mehrotra, 2011). In more than half of the referrals studied there was no communication between the doctors, and between a quarter and half of the doctors who referred did not know whether the patient had been seen by the specialist (Mehrotra, 2011).

The third failure is in expectations. Primary care doctors and specialists often disagree about the specialist's role: a single consultation for an opinion or shared follow-up (Mehrotra, 2011). If the letter does not say what is expected, each side assumes something different.

A review of 22 qualitative studies on the views of family doctors grouped the barriers to communication with specialists into four types: structural, regulatory and procedural, technological, and personal or interpersonal, including lack of trust and mutual respect (Fattahi, 2025). The problem, therefore, is not just about forms.

On the patient's side, the failure shows up differently. The person leaves the consultation with the feeling that something needs to be investigated, but without knowing exactly what or why. They wait weeks or months, arrive at the specialist, and have to tell the story again, often without remembering the tests done or the treatments tried. When they return, they bring a partial explanation and a new prescription. Every poorly handled handoff adds uncertainty for the person who is ill, and trust in the health system wears down along with it.

There is also a cost to the system itself. Tests repeated because the previous result was not sent, consultations spent reconstructing the history, and unnecessary return visits take up slots that could go to those waiting. A well-written letter is, in this sense, a form of care for other patients too.

The filter role of primary care

In many health systems, primary care works as the entry point: access to a specialist depends on a referral. This gatekeeping role was discussed in an analysis that asked whether it would be an effective filter or a failed experiment (Forrest, 2003). The author argued that the quality of the information handover deserves to be studied and standardized (Forrest, 2003).

A systematic review of 25 studies linked the primary care filter to better quality of care, more appropriate referrals and fewer hospital admissions, but it recorded concerns about diagnostic accuracy and one unfavorable study on cancer (Sripa, 2019). Another review, with 26 studies, found lower use of services and lower spending, with limited quality of evidence (Velasco Garrido, 2011).

What these data suggest is that the filter works better when the handover works. A survey of 7,183 family physicians in 34 countries showed that referral letters are sent more often, and replies received more often, where the family physician has a gatekeeping role, where there is secretarial or nursing support, where there is health information technology and where job satisfaction is higher (Scaioli, 2020). Communication depends on structure, not only on good will.

Deciding whether and when to refer

Before the letter comes the decision. Referring is a course of action like any other and deserves the same question asked before ordering a test: what do I expect to change? The text on the question that decides applies here too.

In practice, a referral usually has one of these goals: to confirm or rule out a diagnosis, to perform a procedure that is not available in primary care, to obtain an opinion on treatment or to transfer the follow-up of a complex condition. Each goal calls for a different letter and a different expectation of a reply.

It is also worth looking at your own pattern. Because the referral threshold varies widely between physicians (Mehrotra, 2011), it is useful to review your own referrals from time to time: how many went to each specialty, how many came back with a reply that changed management and how many could have been resolved another way. This review, done in a group with colleagues, is one of the forms of peer feedback that research links to better letters (Tobin-Schnittger, 2018).

It is also worth talking with the person before deciding. Sometimes they ask for a specialist because they fear a specific disease, and a clear explanation settles the concern. Other times they do not want to be referred, because of fear, cost or distance, and that resistance needs to enter the decision. A referral the person does not understand or accept tends not to happen.

When the doubt is small and specific, alternatives to an in-person referral can resolve it. A review of 72 studies on electronic consultation between family physician and hospital specialist linked this tool to better access and to avoided referrals, although the quality of the evidence is low (Peeters, 2024). A well-formulated question, answered in writing, sometimes saves a wait of months.

Urgency also needs to be decided, not left for the specialist to guess. In a study of 264 patients referred to specialized mental health services, triage based on the letter agreed with triage done after the consultation for 143 patients and underestimated urgency for 70 (Nymoen, 2022). The overall quality of the letter did not explain the differences, but low-quality letters were more frequent among cases with underestimated urgency (Nymoen, 2022).

When the reasoning is still open, it is worth saying so in the letter. The text on strategies for deciding without certainty recalls that making the doubt explicit is better than hiding it behind a provisional diagnosis presented as certain.

What the letter needs to say

The analysis of referrals describes the ideal content of a letter: the reason for the referral, the investigation done up to that point and the list of medications in use (Mehrotra, 2011). Other sources add elements that help the specialist decide the priority and the type of care.

A useful letter usually answers these questions:

• what is the clinical question, written directly;
• what are the symptoms, how long have they been present and how have they evolved;
• which tests have already been done and with what result;
• which treatments were tried and what was the response;
• what are the other health conditions and the medications in use;
• what is the degree of urgency and why;
• what is expected from the specialist: opinion, procedure or follow-up;
• what the patient knows and what they expect from the referral.

The clinical question is the center. A letter that says assessment informs nothing. A letter that says chest pain on exertion for three weeks, stress test not available here, I request assessment of ischemia and guidance on treatment allows the specialist to prepare and respond to what was asked.

The summary also matters. Copying the entire medical record into the letter seems complete, but it forces the specialist to look for the relevant information among many others. Ideally the letter can be read in one minute and the specialist knows, on finishing, why the person is there, what is already known and what is expected of them. Additional details can follow separately, as supporting documents.

The family physician's reasoning is a valuable part of the letter. Stating the main hypothesis, which diagnoses were considered and ruled out and why the doubt persists helps the specialist avoid repeating the path already taken. When the family physician suspects something serious, this must be written clearly, and not hidden in the middle of the text.

What the person expects also counts. They may be afraid of a specific disease, or expect a procedure that the family physician does not consider indicated. Recording this avoids frustration and helps the specialist conduct the consultation. When there is a language barrier, the text on consultations with an interpreter reminds us that this information should also be included.

Element	Why it matters	Common mistake
Clinical question	Defines what the specialist should answer	Writing only the name of the organ or the specialty
Summarized history	Avoids restarting the investigation from scratch	Copying the entire medical record without a summary
Tests done	Avoids repetition and shows the reasoning	Citing tests without result or without date
Treatments and medications	Shows what has already been tried and avoids interactions	Omitting medications from other professionals
Urgency	Guides triage and waiting time	Marking everything as urgent or marking nothing
Expectation about the specialist's role	Avoids disagreement about a single consultation or follow-up	Leaving it implicit

Templates and checklists

Writing a letter well takes time, and time is scarce. That is why several interventions have tried to structure the content. A review of 18 publications on improving referral letters concluded that structured templates increase documentation, that peer feedback improves quality and can reduce inappropriate referrals by up to half, and that packages combining several strategies with peer feedback work better (Tobin-Schnittger, 2018). The same review noted that electronic referral brings advantages, but also imposes new barriers (Tobin-Schnittger, 2018).

A randomized study with simulated cases compared free-text letters with letters that included an electronic checklist. Thirty-two gastroenterologists rated the quality of the letters on a scale from zero to ten: the mean was 6.5 for free text and 7.5 with the checklist (Eskeland, 2018). Triage did not change, but fewer specialists needed to request additional information (Eskeland, 2018).

These results point to a balance. The template ensures that nothing essential is left out, but it does not replace the summary. A checklist filled in mechanically can be as unhelpful as a loose line. Ideally there is a short template, with fields for the question, the urgency and what has already been done, and space for the physician's reasoning.

Models also need to fit the routine. A form that is too long tends to be filled in halfway or with generic answers. Required fields for the clinical question and the urgency, with the rest optional, usually work better than an extensive list. The same logic applies to electronic referral: it solves problems of illegible handwriting and lost paper, but it can create others, such as poorly configured fields or messages that nobody reads (Tobin-Schnittger, 2018).

The specialist's reply

The referral is only complete when the information comes back. A survey with 259 family physicians and 232 specialists showed dissatisfaction on both sides: fewer than a third of the specialists considered the family physicians' letters good, only about half of the family physicians felt their question was answered, and fewer than a quarter said the specialist's reply arrived in time (Berendsen, 2009). Both groups said they wanted feedback, but they rarely gave it (Berendsen, 2009).

The reply has practical functions. It reports the diagnosis or the conclusion, says what was done and what is still pending, defines who follows up on what, and guides the family physician on signs that justify a new contact. The analysis of referrals describes three types of continuity that depend on this: of information, of management and of relationship (Mehrotra, 2011).

When the reply does not come, the patient returns to primary care with partial explanations, new prescriptions without justification and doubts about the next step. The family physician has to reconstruct what happened from the patient's account, which is a known source of error. The report on diagnostic error includes failure to communicate the diagnosis among the forms of error.

A good reply, therefore, mirrors a good letter: it answers the question asked, says what was found, explains the changes in treatment and makes clear who does what from that point on.

For the specialist, writing that reply is also a way of teaching. Explaining why a hypothesis was ruled out, which sign changed the approach or when the referral could have been avoided helps the family physician decide better next time. This feedback, when it arrives respectfully, is exactly the kind of exchange between peers that improves the quality of referrals (Tobin-Schnittger, 2018).

Following up on the referral

Referring does not end the responsibility of the one who referred. The analysis of referrals highlights tracking the referral as an essential part of the process (Mehrotra, 2011). Knowing whether the patient got the appointment, whether they attended and what the conclusion was is what closes the loop.

A few simple practices help:

• record the referral, the question and the urgency in the medical record;
• agree with the patient on what to do if the appointment is not scheduled;
• keep a list of pending referrals reviewed periodically;
• ask the patient to bring the reply or the summary of the appointment;
• contact the specialist directly when the reply does not arrive.

The person themselves is an important ally in this follow-up. Explaining what the referral is for, what they should take to the appointment and what they should bring back increases the chance that the information circulates. A copy of the letter in the patient's hands can keep the appointment from starting from scratch when the original letter is lost.

Teamwork can share this task. Nursing and the front desk can track the list of pending items, as suggested by the survey that associated support from the front desk or nursing with more communication (Scaioli, 2020). The text on diagnosis as teamwork discusses how this division of tasks works.

While the patient waits, the responsibility continues. If the symptoms worsen, the urgency needs to be reviewed. The text on planned reassessment applies here: the patient should leave knowing which signs justify seeking help before the appointment with the specialist.

When the relationship between the levels fails

Communication between a family doctor and a specialist is also a relationship between people. The review of qualitative studies cited earlier recorded a lack of trust and respect among the personal and interpersonal barriers (Fattahi, 2025). When one side distrusts the other, the letter tends to be defensive and the reply bureaucratic.

Knowing the colleagues you refer to helps. A phone call when in doubt, a conversation about what each side expects and feedback on difficult cases build trust. The review on improving letters showed that peer feedback is among the interventions that work best (Tobin-Schnittger, 2018).

It is also worth recalling what hinders team reasoning in general. The text on what gets in the way of team reasoning shows that hierarchy, haste and a lack of a communication channel produce the same problems within a service and between services.

A simple way to start is to replace anonymity with contact. Knowing who works at the referral outpatient clinic, having a channel for quick questions and discussing cases in regular meetings brings the two sides closer. When the specialist knows the context of primary care, they understand better why certain letters arrive as they do; when the family doctor knows the limitations of the outpatient clinic, they adjust what they ask for.

In telemedicine, referral takes on other nuances. When the initial assessment was remote, the letter must make clear what was and was not examined, as discussed in the text on when to convert to in person.

It is worth remembering that the letter is also read by the patient, who often carries it in hand. Writing clearly, without terms that might alarm without explanation, respects the person and keeps them from arriving at the specialist with doubts that could have been resolved at the original appointment.

Frequently asked questions

What must a referral letter always include?

The reason, the investigation done and the list of medications in use (Mehrotra, 2011). The direct clinical question and the degree of urgency complete the essentials.

Do structured templates help?

Yes. Templates increase documentation and checklists improve the perceived quality of the letter (Tobin-Schnittger, 2018; Eskeland, 2018).

Does electronic consultation replace referral?

In part. It was associated with better access and avoided referrals, with low-quality evidence (Peeters, 2024). It serves for one-off questions, not for cases that need an examination or procedure.

Who is responsible for knowing whether the patient was seen?

The one who refers. Between a quarter and half of doctors did not know whether the patient had been seen by the specialist, and tracking the referral is part of the process (Mehrotra, 2011).

Why must urgency be written down?

Because triage based on the letter can underestimate urgency, especially when the letter is of low quality (Nymoen, 2022).

What should you do when the specialist's reply does not arrive?

Contact the specialist directly and review the list of pending items. Fewer than a quarter of family doctors said they received the reply in time (Berendsen, 2009).

Final checklist

• Before referring, define what you expect to change with the specialist consultation.
• Write the clinical question directly, on the first line of the letter.
• Include symptoms, tests with results and dates, treatments tried and medications in use.
• State the urgency and explain why.
• Say whether you expect an opinion, a procedure or shared follow-up.
• Record the referral and follow it up until you receive the reply.

References

1. Berendsen AJ, Kuiken A, Benneker WH, Meyboom-de Jong B, Voorn TB, Schuling J. How do general practitioners and specialists value their mutual communication? A survey. BMC Health Serv Res. 2009. PMID 19664238. DOI 10.1186/1472-6963-9-143.
2. Eskeland SL, Rueegg CS, Brunborg C, Aabakken L, de Lange T. Electronic checklists improve referral letters in gastroenterology: a randomized vignette survey. Int J Qual Health Care. 2018. PMID 29635304. DOI 10.1093/intqhc/mzy057.
3. Fattahi H, Seproo FG, Fattahi A, Rostami V, Shokri A. General Practitioners' Perspectives on Barriers to Communication With Specialists in the Referral System: A Systematic Review and Meta-Synthesis. Health Sci Rep. 2025. PMID 40415981. DOI 10.1002/hsr2.70785.
4. Forrest CB. Primary care in the United States: primary care gatekeeping and referrals: effective filter or failed experiment?. BMJ. 2003. PMID 12663407. DOI 10.1136/bmj.326.7391.692.
5. Mehrotra A, Forrest CB, Lin CY. Dropping the baton: specialty referrals in the United States. Milbank Q. 2011. PMID 21418312. DOI 10.1111/j.1468-0009.2011.00619.x.
6. Nymoen M, Biringer E, Hetlevik Ø, Thorsen O, Assmus J, Hartveit M. The impact of referral letter quality on timely access to specialised mental health care: a quantitative study of the reliability of patient triage. BMC Health Serv Res. 2022. PMID 35655302. DOI 10.1186/s12913-022-08139-3.
7. Peeters KMM, Reichel LAM, Muris DMJ, Cals JWL. Family Physician-to-Hospital Specialist Electronic Consultation and Access to Hospital Care: A Systematic Review. JAMA Netw Open. 2024. PMID 38214930. DOI 10.1001/jamanetworkopen.2023.51623.
8. Scaioli G, Schäfer WLA, Boerma WGW, Spreeuwenberg PMM, Schellevis FG, Groenewegen PP. Communication between general practitioners and medical specialists in the referral process: a cross-sectional survey in 34 countries. BMC Fam Pract. 2020. PMID 32183771. DOI 10.1186/s12875-020-01124-x.
9. Sripa P, Hayhoe B, Garg P, Majeed A, Greenfield G. Impact of GP gatekeeping on quality of care, and health outcomes, use, and expenditure: a systematic review. Br J Gen Pract. 2019. PMID 30910875. DOI 10.3399/bjgp19X702209.
10. Tobin-Schnittger P, O'Doherty J, O'Connor R, O'Regan A. Improving quality of referral letters from primary to secondary care: a literature review and discussion paper. Prim Health Care Res Dev. 2018. PMID 29212565. DOI 10.1017/S1463423617000755.
11. Velasco Garrido M, Zentner A, Busse R. The effects of gatekeeping: a systematic review of the literature. Scand J Prim Health Care. 2011. PMID 21192758. DOI 10.3109/02813432.2010.537015.