Most paediatric diseases are managed and resolved at the primary care (PC) level without need of diagnostic tests (DTs). However, in some infectious diseases, access to DTs can increase the accuracy of diagnosis, help determine appropriate treatment and make earlier and appropriate referrals.1 Barriers to access to DTs in PC paediatric clinics (ordering test, performance of test, delivery of results, turnaround times) can affect or hinder clinical practice, rational use of antibiotics, performance of research studies and knowledge of local epidemiology and increase the frequency of referral to hospitals, among other aspects.

Primary care paediatric services are located in different settings: urban, semi-urban and rural, in which the accessibility of laboratory, microbiology and radiology services may vary based on distance, hours of operation and frequency of sample collection and the referral or submission pathways established in the different health care areas. Based on previous studies, only 29% of providers can order all laboratory or imaging tests without restrictions from the PC setting, and only 18% have access to all tests following a given protocol.2,3

The objective of the study, promoted by the Group on Infectious Diseases (GID) of the Asociación Española de Pediatría de Atención Primaria (Spanish Association of Primary Care Paediatrics, AEPap),4 was to determine the access of primary care paediatricians (PCPs) working in the public health system to ordering and performing tests for diagnosis of infectious diseases in their health care centres, their access to tests results and the associated turnaround times.

We conducted a cross-sectional, observational descriptive nationwide study in PCPs employed in the public health system through completion of a questionnaire, on a voluntary and unpaid basis, distributed electronically between April and May 2017 (available at https://goo.gl/forms/VYNoqiuQhojACv2Z2). We recruited participants from the list of members of the AEPap, a professional association of PC paediatricians, and also sent the questionnaire to subscribers of the PEDIAP mailing list,5 created in 2000 as a platform for professionals to network and discuss subjects of interest in the field of PC paediatrics, who amount to approximately 1000 and are very representative of the PC paediatrician collective. We specifically requested that only paediatricians involved in PC, exclusively or in combination with other activities, respond to the survey, clarifying that responses should refer solely to the primary care setting. In the 2 months that the questionnaire was available for completion, we sent 2 reminders.

We collected data on the characteristics of the respondents, including age, sex, work setting, caseload, availability of paediatric nursing staff and the degree of access to a hospital. When it came to diagnostic tests, we grouped them by type of disease: acute pharyngotonsillitis (AP), urinalysis (urine strip test, sediment and culture); other bacterial cultures and serological tests, diagnostic tests used in pertussis and tuberculosis and imaging tests.

We expressed and tabulated qualitative data as percentages with the corresponding 95% confidence intervals (CIs). We expressed quantitative data as mean and standard deviation. We performed a univariate analysis with calculation of odds ratios (ORs) to make comparisons, using the  χ2 square test and defining statistical significance as a p-value of 0.05 or less.

Most infectious diseases managed by PCPs do not require performance of diagnostic tests, but the availability of the latter is relevant when it comes to the decision whether to prescribe an antibiotic agent, to obtain data on antimicrobial resistance in the community on which to base recommendations for empiric antibiotherapy and to identify and control the spread of infections. Diagnostic tests can also allow earlier diagnosis and initiation of treatment.

Based on previous data, it appears that the number of days in which samples can be collected and submitted remains stable (2012 PAPE survey)3: 71% of respondents could collect and submit blood samples Monday through Friday.

Only 31.9% had laboratory test results automatically uploaded to patient health records, a proportion that should be increased to integrate information and to facilitate and expedite access to test results by primary care providers.

Acute pharyngotonsillitis is one of the most common childhood diseases, and while there are guidelines for the prescription of antibiotherapy, there is still evidence of frequent and inappropriate prescription for treatment of this disease.6,7 The aetiology is usually viral, and the most frequent bacterial causative agent is Streptococcus pyogenes or group A beta-haemolytic streptococcus ( (GABHS). Even in patients with the highest scores in the most widely used clinical rules, the probability of APT caused by S. pyogenes does not exceed 60% to 64%, so the suspected diagnosis should be confirmed by microbiological methods. Although throat culture is the gold standard, a high percentage of paediatricians did not use it routinely in their practices due to the long turnaround times. There is evidence of the usefulness of RADTs,8 possibly exceeding that of culture, in reducing antibiotic prescription in the context of APT,9 and therefore these tests should be used by every provider. At present, while they are widely used (79% of respondents reported ordering these tests), the reason given by 95% of those that did not use rapid tests was that they were not available, with substantial differences between autonomous communities in their availability.

Collection of samples for bacterial culture has proven useful10 to improve rational prescription of antibiotics at the time of initiation of treatment and for its adjustment, suspension or replacement by a different agent once results are available. Restrictions in the times that samples can be collected or submitted at the primary care level (1/3 of respondents could not collect samples on every weekday, and the time window for submission was 1 h for nearly half of providers), and delays in obtaining the results (≥ 4 days for 44% of respondents for urine cultures and for 50% of respondents for throat cultures) are a significant barrier to rational use of antibiotics. This may entail referral of patients to hospital-based services (which are difficult to access for 43% of respondents), in which case not only may patients be inconvenienced by having to travel to the site, but the performance of the tests requested by the PCP depends on decisions made by a different professional, which thwarts the considerable capacity of primary care services for resolving problems presenting at that level.

These limitations particularly affect the diagnosis and treatment of urinary tract infections (UTIs), in which confirmation of the diagnosis, identification of the causative agent and establishment of antimicrobial resistance patterns are essential to provide appropriate treatment to the patient and obtain information on the local epidemiology. The diagnosis of UTIs continues to be a challenge, especially in children without bladder control, due to the nonspecificity of the symptoms and the difficulty of obtaining a valid sample. In these children, most samples are obtained with a collection bag or through a midstream catch, methods associated with a high frequency of contamination.11 Thus, positive culture results these samples require confirmation in catheter urine sample, and therefore referral of the patient for performance of this technique (only 18% performed insertion of urinary catheters at the health care centre). A recently published guideline on the management of UTIs excludes the option of performing cultures of samples obtained with a collection bag,12 adherence to which would entail the referral of nearly all patients with suspected UTI.

On the subject of the turnaround time, a study conducted in Andalusia in 200913 found that 3% of PCPs received the results of complete blood counts within 24 h compared to 83.3% of hospital-based paediatricians (HPs). A similar pattern was found for X-ray results, received in 1 day by 68.8% of paediatricians in primary care compared to 85.7% of hospital paediatricians. The differences in the availability of urine culture results at 48 h were even more marked, corresponding to 4.5% of PCPs compared to 45.2% of HPs, with significantly longer turnaround times at the primary care level.

In spite of these factors, the collection of samples for culture in PC is relevant for the purpose of establishing the epidemiology of community-acquired infections, the causative agents involved and their drug resistance profile.14,15

Access to DTs and a short turnaround time are particularly relevant in diseases like tuberculosis (TB) and pertussis, in which early diagnosis is key to initiate treatment and control outbreaks. The Mantoux test is the gold standard for initial diagnosis and contact screening of TB.16,17 It is usually performed in the primary care setting, although our findings show the effects of a temporary shortage of tuberculin antigen. As for pertussis, since neither the infection nor vaccination confer lifelong immunity and the presentation is often atypical, a high level of suspicion is required for early detection of cases and prevention of transmission,18 while a certain diagnosis is important to prevent the use of antibiotics in unconfirmed cases. At present, PCR is considered the gold standard19 and it was available to 87% of respondents, although 62% of PCPs needed to refer the patient for sample collection, with differences between autonomous communities. Once again, the long turnaround times (only 2% receive the results in the day and 23% in 1–2 days) pose a barrier to early diagnosis, treatment and initiation of antibiotic prophylaxis.

In general, serologic tests were ordered infrequently at the primary care level, perhaps because they tend to be ordered to guide clinical decision-making and the delays in obtaining the results undermine this purpose.

The results involving imaging tests were pretty homogeneous across autonomous communities. Access to these tests was optimal, although the delays in getting an appointment for elective ultrasound examinations could be improved.

When it came to ultrasound examinations, primary care paediatricians had little autonomy, as the actual performance of requested tests was left to the judgment of hospital-based providers.

There is evidence that point-of-care ultrasound performed by clinicians, including primary care paediatricians,20 may be useful for diagnosis of different diseases.21,22 However, few paediatricians order or perform point-of-care ultrasound examinations at the primary care level due to their excessive workloads and the lack of sonographers on staff.23

The task of collecting samples for testing falls within the nursing scope of practice, but we found a considerable involvement of paediatricians, independently of the size of their caseloads. This may be explained by the frequent lack of paediatric nurses in primary care settings, as many operate under a community health nursing model in which nurses are shared by paediatric and family medicine teams, have greater workloads and are less familiar with paediatric care, and PCPs end up performing these tasks.24

When we compared the turnaround of results in sites serving populations greater and smaller than 10 000 inhabitants, we found that the results were not delayed in smaller towns.

There are no studies in the literature providing data on the overall access of primary care paediatricians to the ordering and/or performance of diagnostic tests. It would be interesting to replicate this study in the future to analyse how the availability of diagnostic tests changes over time.

We found differences between autonomous communities and care settings. There are clear opportunities for improvement in the access to different DTs, the time elapsed to collection and submission of samples, the test turnaround time and the wait times for performance of elective imaging tests. This interferes with the ability of primary care paediatricians to manage and resolve presenting complaints.

The authors have no conflicts of interest to declare.