In oncological patients, the interaction of multiple factors related to immune supression, skin or mucosal disruption, extensive antibiotic use, complex surgical procedures and the presence of invasive devices, augment the risk of HAIs, leading to an increase in the length of hospital stay, morbidity and mortality [13].
As a result, the number of patients admitted to the ICU, either for cancer-related complications or for treatment-associated side effects is steadily increasing [9].
Although we did not specify the chemotherapy regimens, it is important to consider that hematologic patients, particularly with acute leukaemias, aggressive regimens amplify disease-induced immunosupression and result in severe and long lasting neutropenias which could favour the development of severe infections. This was not confirmed in this study, most likely because of the low number of patients with this disease.
Some studies describe as the main reasons to the ICU admission, septic shock, respiratory failure and hypovolemic shock secondary to extensive surgical resection in patients with solid tumors. In this study we found differences for ICU admission in both groups: in patients with HM the most frequent cause was respiratory failure (54.7 %), it was related to several causes such as pneumonia, acute respiratory distress syndrome, diffuse alveolar hemorrhage and pulmonary embolisms. However, in patients with solid tumors the main reason for ICU admission was hypovolemic shock secondary to schedule or emergency surgical resection (29.9 %). This percentage was higher than the one reported in other study performed in the same ICU (31.6 % for septic and hypovolemic shock) [14]. Reports from other parts of the world not even include hypovolemic shock within the first causes of admission to ICU [15–17]. One possible explanation for the high percentage we found, is that patients admitted to our hospital have advanced cancer stages, with tumor adhesion to adjacent organs, therefore require extensive surgeries, with multiple organ resection, which increases bleeds and surgical morbidity.
HAI prevalence varies among hospitals and among countries. There has been reported an increase in patients with cancer during the last decade [18]. Some studies have reported a wide variation rate between 5.3 and 56.1 % [19]. Our HAI rate was 40.7 %. The most frequent infection was VAP (26.7 %, 28.4/1,000 ventilator-days). The VAP rate was similar to the ones described in other ICUs in developing countries (24.1/1,000 ventilator-days) [20]. More than five days of mechanical ventilation was the only risk factor associated to HAI in multivariate analysis (OR 3.12, CI 95 % 1.6 – 6.2, p?=?0.001), as has been reported previously [21].
Surgical site infection (SSI) was the second HAI documented: 72 surgeries were performed during the study period; the infection rate was 16.7 (the SSI infection rate reported during 2013 in our institution was 9.9). It is relevant to highlight that 28 % of the surgeries were emergency procedures, and the majority of these comprised complicated and extensive interventions.
In various reports, CA-UTIs are described as the most common nosocomial infection (around 40 %) [22]. These infections do not cause severe mortality or morbidity, but can progress into severe infections such as secondary bacteremia or septic shock, in addition to significantly increasing the hospital costs [23]. We found CA-UTI as the third HAI in the ICU (13.4 %, 8.9 infections/1,000 catheter-days), as the same as reported by other authors [20]. Median of indwelling urinary catheter stay were 16 days in patients with CA-UTI compared with 10 days in non-CA-UTI patients.
The CLABSI rate was 2.5 % (1 infection/1,000 catheter-days). In some U.S. studies, which report prevalence, the range varies from 1.8 to 7.6/1,000 catheter-days [24]. In one research conducted in eight developing countries, the rate varied between 4.2 and 14.4/1,000 catheter-days [17]. The low rate found in this series is attributed to strict standardized protocol of care provided by an intravenous therapy team nurses working for over two decades [25, 26].
E. coli was the most frequent pathogen isolated, 78 % were ESBL producers. This percentage is considerably higher than reported elsewhere in our country (range between 33 and 51 %) [27–29], but it is important to note that these studies included patients hospitalized in different wards, not only from ICU where the prevalence of MDRB is usually higher.
In recent years, widespread MDRB in hospitals had complicated the medical care and infection prevention measures. MDRB pathogens are associated with worse outcomes related with inferior therapeutic options and delay in initiation of appropriate therapy, and probably with increased virulence of these strains [30]. In this study we found 24.2 % of MDRB, which is lower when compared to a retrospective study performed from 2007 to 2011 in the same ICU, when the HAI rate was 39.7 % [31]. In another Mexican study with patients admitted in ICU, the incidence of MDRB was 64.5 % [32].
In this report, overall 30-day mortality was 39 % with ranges between 28 % in patients with solid tumors until 60.4 % in HM group. The poor prognosis of patients with HM, especially those who received hematopoietic stem cell transplantation (HSCT), has been well documented [17]. We did not have HSCT patients because they are treated exclusively in the transplant unit. We found a difference in mortality related with the cause of admission to ICU: 60 % in patients admitted with respiratory failure, 29.4 % in patients with septic shock and 10.3 % in scheduled surgical patients, but this could be related with the oncology disease, because HM patients had also more frequently respiratory failure. In another study, that included oncology patients admitted at the ICU, the overall hospital mortality was 30 %, (range, 10 % to 70 %) [9, 17].
Early management of organ dysfunction is crucial in the treatment of critically ill patients; late ICU admission has been associated with greater mortality [9]. We found a significant delay in ICU admission in patients with HM (median 2 days compared solid tumors median 0 days), which could be also related with higher mortality in these patients.
Infection control measures are strict in our ICU: at admission, two blood cultures, urine, respiratory secretions and other cultures are required; broad-spectrum antibiotics are initiated as soon as patient arrival. An antimicrobial re-evaluation is made in the first 48–72 h when cultures results are available and according to clinical evolution to either de-escalate or escalate the antibiotic prescription; antibiotics schemes are discussed and reviewed periodically with all staff (intensive care, infectious disease and medical or surgical oncologists physicians); the duration of antimicrobials treatment are shortened as much as possible; surveillance hand hygiene program, and follow antibiotic prophylaxis guides [2].
Some limitations of this study are that the results correspond to a single-center and did not obtain data related to the long-term survival. However, we believe these results provide insight into the prevalence of HAI, particularly MDRB at a cancer referral center, and can be extrapolate results to other hospitals with similar characteristics.