Ananalysis ofdecisions byEuropeangeneral practitioners toadmit tohospital patientswith lower respiratory tractinfections

Background -Thepurposeofthisstudy wastoidentify factors onwhichEuropean general practitioners (GPs)basetheir decisions toadmittohospital patients withlowerrespiratory tractinfections (LRTI). Methods-A surveywascarried outfrom December1993toJanuary1994toidentify factors thataffect GPs'decisions toadmit tohospital patients withLRTIbycollectingdataon2056patients from605GPsin France,Germany,Italy, Spain,andthe UK. Results - Only93(4.5%) ofthepatients included inthestudywereadmitted to hospital. Univariate analysis showedthat age> 60years, institutionalisation ofthe patient, concomitant diseases, cardiac insufficiency, asthma, adiagnosis ofpneumonia,andclinical signssuchaschest pain,cyanosis, tachypnoea andhypotensionsignificantly (oddsratio(OR)> 2.0, p < 0.002)influenced the decision to admittohospital. No influence couldbe shownforsex,smokinghabits, history of bronchiectasis orchronic bronchitis, the presence offever, chills, myalgia, coughor purulent sputum,andthediagnoses of acutebronchitis, influenza orexacerbationofchronic bronchitis. Inthemultivariate analysis onlythepresence ofchest pain(OR2.3,95% confidence interval (CI) 1.5to3.5), cyanosis (OR4.1, 95%CI2.4to 7.1), dyspnoea (OR4.9,95%CI3.1to7.9), andhypotension (OR2.9,95%CI1.6to 5.2), aswellasadiagnosis ofpneumonia (OR6.6, 95%CI4.3to10)(all p < 0.00001) remainedas factorsthatsignificantly affected thedecision toadmittohospital. Conclusions - Clinical signsofsevere infection anda diagnosis ofpneumonia arethemainfactors thatinduceGPsto admitpatients withLRTItohospital in Europe. (Thorax 1996;51:1017-1022)