Background and aim In order to provide patients with the best care, physicians’ choices should follow guidelines that are constructed based on the best reliable data in order to increase the probability of success. This is the principle of Evidence Based Medicine (EBM). Blood tests are prescribed everyday by family doctors and these must be chosen criteriously to prevent overdiagnosis, overtreatment and unnecessary costs.This study was designed to understand if the prescription of blood analyses in Family Medicine in Portugal is done according to EBM.

Methods Firstly, we performed a restrictive systematic review in order to define the evidence based reccomendations for the prescription of each blood test.

Posteriorly, we conducted a cross-sectional study having as setting the Portuguese primary healthcare. Doctors working in primary care in Portugal were surveyed, through an online questionnaire, between January and July of 2017. The questions were based on the panel of the most prescribed analyses in the group of healthcare centres of Central Lisbon in order to evaluate the motives for prescribing or not prescribing blood tests and if this prescription is performed according to EBM. The statistical analysis of the collected data was performed using the software IBM® SPSS® Statistics 23.

Results There were 323 valid answers. Most of the doctors affirmed to prescribe blood tests according to EBM (64%). A statistically significant relation between this affirmation and the motives chosen for the prescription of each blood test was verified (full blood count (fbc) p<0.001, urinalysis p<0.001, lipid profile p<0.001, glycaemia p=0.002, liver function p=0.036, renal function p=0.005, thyroid function p=0.026, sedimentation rate p=0.032 ) except regarding the prescription of prostate specific antigen (PSA) where most doctors, even the ones that did not affirm to following EBM recommendations, are prescribing correctly. Many doctors affirm to prescribe blood tests to asymptomatic, risk free patients (fbc (48%), glycaemia(57.6%), lipid profile (67.2%), renal function (39%) ). For the prescription of most blood tests, signs and symptoms of a disease was the most chosen motive (fbc (67.2), glyceamia (84.5%), liver function (82.7%), renal function (80.2%), thyroid function (93.8%), sedimentation rate (60.1%), PSA (77.4%)), except for the prescription of lipid profile for which personal history of cardiovascular disease was the more common answer (94.4%).

Discussion and conclusion There are still many prescriptions of blood tests that do not follow the recommendations, like the prescription of full blood count to screen asymptomatic, risk free patients.Despite this, we can verifiy that doctors that are familiar with and affirm to base themselves in EBM tend to follow criterious guidelines for the prescription of each blood test. The exception is the prescription of PSA, probably for being a widely discussed topic and doctors that are not familiar with EBM also know the recommendations.

This study allows us to understand that there is a significative relation between knowing and applying EBM and that it is still necessary to disseminate the concept of EBM in the practice of Family Medicine in Portugal in order to avoid unnecessary and maybe harmful prescriptions of blood tests.