An analysis of referrals to General Pulmonology consultation is extremely pertinent in today's world. There is increased demand and the clinical information available is at times limited, meaning that the whole process could be optimised, which could potentially impact on respective waiting periods.

The standard and pace of referral varies and depends on the individual patient, the nature of the physician's approach and the organisational system concerned, the access to hospital specialists and the relevance of the case (1, 2). Around 5% of patients seen in General Practice (GP) are referred to secondary care.2 The act of referral is not a simple, mechanical process, but a complex interaction, in which the GP should know what is required or expected for that particular referral.1

Studies developed in the area of referrals have shown that there is great potential for improving quality in this context.1 Over-referral has costs and can expose patients to unnecessary risks; under-referral may deprive them of possible beneficial treatment.3 The aim of referral management is to reduce the number of requests as well as influence the endpoint or improve the characteristics of the referral.4

We conducted an analysis of all referrals made by GPs from Primary Care Providers (PCPs) of the central region of Portugal, for General Pulmonology consultation at the Centro Hospitalar e Universitário de Coimbra - Hospitais da Universidade de Coimbra, between January 1st and December 31st 2016. For each referral there was an evaluation of demographic data, the time between referral and appointment, and the situation of the patient on the date these data was collected. Research included: signs/symptoms, physical examination, pulmonary auscultation, personal and family history, regular medication, allergies, environmental/professional exposure, smoking habits, respiratory care, complementary diagnostic tests (CDTs) and the diagnosis which led to the referral (confirmed or suspected).

There were 308 referrals (average 1.2 referrals per working day) from 38 distinct institutions (Personalised Healthcare Units – UCSP and USF – Family Health Units), with an average waiting time between the request and the initial appointment of 64 days (Table 1).

The results are presented in Table 2.

In terms of follow-up of patient care up to the date of this analysis, 41.2% (n=127) continued to attend consultations, 7.2% (n=22) were referred to a sub-specialist consultation and 1.3% (n=4) died. A total of 18.2% (56) failed to attend consultation, 11% (n=34) the initial one and 7.2% (n=22) follow-up. The remaining 32.1% (n=99) were discharged, on average 4.3 months after the start of the follow-up; 29.3% (n=29) were discharged shortly after their initial consultation.

Analysing these results we can conclude that for fewer than 20% of requests, there was a description of the main complaint which motivated the referral along with an objective examination or pulmonary auscultation – foundation blocks in patient assessment, and that, up to the date of consultation, only around 24% had had an imaging test, together with a pulmonary function test. Curiously, 50% of requests included the patient's personal history and 35% their regular medication use, a fact that may be associated with being able to transcribe this information on the referral system. Only 32% of referrals mentioned the known or suspected diagnosis. It has already been reported that the reasons for referral, socio-psychological factors or follow-up plans, as well as clinical findings, test results or previous treatment were often missing.5

It is worth noting that 18% of patients missed initial or follow-up appointments, a fact which can be linked to travel-related or economic difficulties, barriers to mobility or patients not taking responsibility for their state of health. It should be noted that around 32% of patients were discharged on average 4.3 months after the initial consultation, while approximately one third of these were discharged after the initial consultation. This raises questions about the pertinence of some of the referrals.

Some studies suggest that a combination of peer revision, the use of structured protocols and feedback from the specialist could be effective in reducing the referral ratio.4 Strategies aimed at changing referral behaviour by GPs may be effective both in reducing these, and improving the characteristics of the referrals.3

In our unit the existing referral system does not allow requests for additional information about the referred patient and so there are no refusals based on lack of clinical data. Nevertheless it is important to reflect that referral procedures which are properly informative and directed, may lead to the improved screening of submitted requests, optimising existing hospital resources. Taking this into account, models or protocols to this effect may be useful, facilitating the process, and providing an interaction with the requester in order to improve a good referral assessment, minimising the lack of essential information, and enriching our patient safety.

The purpose of protocols is to obtain all necessary information based on “obligatory items” for each parameter so that urgency of the requests is better stratified, time to precise diagnose improved and more appropriate treatment achieved. This will contribute to a safer and faster patient approach.

Therefore, this strategy appears to be something which needs to be developed and implemented, with a subsequent evaluation of its effect and safety, in order to estimate the impact of the measures taken for all the elements involved.