Performing diagnostics, especially blood cultures, on-time for infectious patients reduces length of stay and costs

Objectives As an academic center, the University Medical Center Groningen in the Netherlands receives large numbers of patients with (complex) infectious problems. It is imperative that the correct diagnostics are performed on time for these patients in order to correctly diagnose them and treat them in the most optimal way. This improves quality of care and reduces chances for resistance development due to unnecessary broad spectrum use of antibiotics. The precise effects of these diagnostics are however difficult to quantify. We therefore tried to provide more clarity on this particular subject. Methods Data was collected retrospectively from three different databases within the hospital and combined for analysis. Data covered one whole year (2013). To investigate patients coming to the hospital with (suspected) infectious problems, the data was filtered for patients receiving at least 3 days of broad spectrum antibiotics starting from admission, excluding haematological patients and children (<18). For this filtered group different outcome measures were evaluated: performed clinical chemical diagnostics on day of admission; performed microbiological diagnostics within a timeframe of 3 days of admission; length of stay (LOS); total amount of Daily Defined Doses and total costs of performed procedures. Results In 2013 a total of 785 patients were admitted to the hospital and were prescribed at least three days of antibiotic therapy starting from the admission day or the day after admission. For 205 patients (26%) there were no blood cultures taken during the admission day (+/- 1 day). Mortality between these two groups was similar (3% versus 4%; p=0.672), however LOS was significantly longer for the group that did not receive blood cultures within this timeframe (13.3 versus 16.6; p<0.001). Other clinical chemical diagnostics such as CRP, leucocytes, creatinine and Hb were also less frequently performed within this group. Total costs, based on all performed procedures were significantly lower per patient (€10,128 versus €14,960; p<0.001). When looking exclusively at patients with an ICD9 code for an infectious problem, the effects are even stronger. Conclusions Assuming that patients who receive at least 3 days of broad spectrum antibiotic therapy at admission have a (severe) infectious problem, we could investigate the effects of on-time diagnostics. Performing these diagnostics during admission can reduce LOS with 20%, probably because treatment can be tailored to the patient’s specific needs. The drawing of blood cultures is an essential component of this process and should neither be overlooked nor cut back. The investments on these diagnostics are giving a significant return due to a reduced LOS. Although appropriate protocols do exist, they are apparently not followed correctly. Our data shows that correct diagnostics improve the quality of care and more education and audits are thus needed to improve compliance.