Surveys have shown that HA-BSIs were responsible for 20%-60% of hospitalization-related deaths (Luzzaro et al., 2011; Ziegler et al., 2014). Despite the high morbidity and fatality, to our knowledge, there are few published reports on HA-BSI prevalence estimates and associated mortality in Arab countries, including Oman. This study embarked to explore efforts towards reducing mortality associated with HA-BSI. The study provided supportive evidence for a varying occurrence in HA-BSI case fatality rate estimates over selected socio-demographic and clinical characteristics.
The first aim of this study was to explore whether there is a difference in the socio-demographic characteristics between deceased and survived HA-BSI patients. According to our study, deceased HA-BSI cases were significantly older than survived. Regarding the age groups, Omani governorates, and year of admission, there were no differences among both study groups. Advanced age is considered a factor predisposing to mortality in patients with HA-BSIs. A Belgian study reported old age as an independent risk factor for mortality among patients with HA-BSI where (HR = 1.7; 95% CI: 1.2, 2.4, P = 0.002) (Reunes et al., 2011). According to the EUROBACT international cohort study, significantly higher 28-day mortality was reported among older HA-BSI patients admitted in the ICU (Alive 57.2 + 18.3 vs. Dead 63 + 15.8, P < 0.0001) (Tabah et al., 2012). A study by Keith found that BSI was associated with more than two times greater mortality in older adults (Kaye et al., 2014). In a survey in France, including 121 hospitals, it was discovered that 10-day mortality related to BSI is associated with age over 72 years (P < 0.0001) (Robineau et al., 2018). In Belgium, Blot concluded that in-hospital mortality significantly increased with age: 42.9% in middle-aged patients, 49.1% in old patients, and 56% in very old patients (P < 0.015) (Blot et al., 2009). The risk of one-month mortality from CA-BSI was greater among elderly patients with a median age of 75 years (RR = 5.2; 95% CI: 2.1–12.6). The patients who died after one month of HCA-BSI were with median age of 65.5 years (Raymond et al., 2006). The study by Kontula et al. mentioned that old age was significantly responsible for 28 days of mortality among patients with HA-BSI, with a median age of (67 dead vs. 59 survivals) (Kontula et al., 2018).
In our study, the male gender was significantly more observed among deceased HA-BSI cases than the survived HA-BSI cases. Studies revealed that sex identity significantly contributes to the outcome of HA-BSIs. In a study conducted at Beilinson university hospital, it was found that women with HA-BSI had about a one-third greater risk of death from HA-BSI than men (OR = 1.7; 95% CI: 1.1, 2.6), even when adjusting for other risk factors, where the death rate among women was 43% and 33% among men (P = 0.0001) (Leibovici et al., 2001; Apostolopoulou et al., 2014). In Sweden, Martin at. al stated that the male sex is a risk factor for 30-day mortality (RR = 1.05; 95% CI: 1.01, 1.11) (Holmbom et al., 2016).
Secondly, the study estimated the overall HA-BSI observed and standardized CFR of 28.1%, found similar to CFR obtained in Thailand (28.3%) and comparable to results from the United States (27%) (Edmond et al., 1999; Wenzel & Edmond, 2001) and some European countries like Spain (27.3%) (Vallés et al., 2008) and Estonia (31%) (Mitt et al., 2009). This rate was lower than the rates reported by Israel (Leibovici et al., 2001), Brazil (Marra et al., 2011), and Italy (Orsi et al., 2002) (37.3%, 40% and 57.1%, respectively) and much higher than those reported in Scandinavian countries where Finland (14%) (Kontula et al., 2018) and Sweden (17.2%) (Holmbom et al., 2016) and other European countries where the CFR in Ireland (Brady et al., 2017) is 15.3% and same rate reported in Greek (Starakis et al., 2010). Much lower CFR was reported in Australia (7.1%) (Barnett et al., 2013).
The third aim of our study was to estimate the observed and standardized CFRs of HA-BSI among hospital admissions in association with their socio-demographics. The gender-specific observed and standardized CFRs of HA-BSI were higher among males than females (24.5% and 18.6% vs. 31.2% and 43.2%). Regarding the spectrum of age categories, our study found that the CFR was high among young ages 0–15 (26.9%) and then decreased, reaching the lowest value of 12.9% at middle age (26–35) years. The pattern of the line after the age of 26 years illustrates a direct proportion between standardized CFRs and age. This finding is consistent with other studies findings, as age is a significant risk factor for HA-BSI-associated mortality (Robineau et al., 2018; Blot et al., 2009; Kaye et al., 2014).
We found that the observed CFR was highest among patients from the Dhahira governorate (30.4%). Dhofar came next, followed by Muscat, Sharqiya, Batinah, and Dakhliya governorates. The lowest observed CFR was among Buraimi HA-BSI patients (18.2%). The results showed that the regional-specific standardized CFR was highest among HA-BSI patients from Dakhliya (42.5%), followed by Dhofar, Dhahira, Sharqiya, Muscat, and Batinah. The lowest standardized CFR was among HA-BSI patients from Buraimi (11.5%).
Over the study period from 2015 to 2019, the highest observed case fatality rate was 29.2%, reported in 2017. The highest standardized CFR was 31.8% in 2015, and the lowest standardized CFR was 24.3% in 2019.
The results showed that the CFR was highest (52.6%) among HA-BSI patients admitted in the ICU compared to other hospital wards. This finding logically corresponds to the fact that BSI related to ICU admission is significantly associated with high in-hospital death rates among ICU admissions (Prowle et al., 2011; Kontula et al., 2018). Our observed CFR rate among ICU admissions was comparable to that reported in a parallel study, where 50.4% was the CFR among ICU admissions with HA-BSI (Leibovici et al., 2001). Another study detected lower CFR among HA-BSI ICU admissions (35.3%) and indicated no association between HA-BSI and increased ICU mortality (Giovine et al., 1999). According to a study in Finland, 28 days of death is significantly frequent in patients with BSI related to ICU admission (P < 0.001) (Kontula et al., 2018). BSI related to ICU admission is associated with high in-hospital death rates among ICU admissions, 41.2% (Prowle et al., 2011). The 28-day mortality rate is higher among HA-BSI patients with an abdominal source of infection, as detected by EURPBACT international cohort study (Tabah et al., 2012). In France, digestive tract infection and pneumonia as a source of BSI were associated with mortality (P < 0.0001) (Robineau et al., 2018). In a study from New Zealand, mortality was associated with catheter site BSI as the common source compared to the other sources combined (P = 0.04). Early (3–7 days) mortality in patients with HA-BSI was significantly frequent in those with catheter-related BSI and chronic hemodialysis as a source of BSI. In patients with secondary BSI, the respiratory tract (24%), gastrointestinal tract (21%), and urinary tract (21%) were the primary sources of infections noted in patients who died within two days. Patients who survived more than 28 days were most frequently among patients with BSI related to admission for obstetric care and newborns (Kontula et al., 2018).
Our study reported that the highest CFRs were among HA-BSI patients with liver disease, pulmonary disorders, and cancer, with 62.5, 51.5 and 44.9%, respectively. This finding agrees with the study that found HA-BSI patients with liver cirrhosis are 17 times more susceptible to death (Hortiwakul et al., 2012). Significantly high mortality rates were observed in HA-BSI patients with liver diseases (52.3%) and respiratory diseases (49%) (Tabah et al., 2012), Thus consistent with our study findings. In a Finnish study, solid malignancies were associated with high CFR related to HA-BSI (Kontula et al., 2018). The underlying disease and comorbidities may complicate the outcome of the BSI. EURPBACT international cohort study reported a significant increase in the 28-day death rate among HA-BSI patients with chronic respiratory disease and immune deficiency in ICU (Tabah et al., 2012). In the French study, the mortality rates were significantly lower among patients with renal failure (P < 0.0001) (Robineau et al., 2018). In New Zealand, mortality rates were highest among adult BSI patients with hematology-oncology (Raymond et al., 2006). Solid Malignancies are significantly reported in patients who died within 28 days of HA-BSI infections (P < 0.001), according to the Kontula study (Kontula et al., 2018). Increased mortality was independently associated with liver cirrhosis (OR = 17.5; 95% CI: 2.4, 125.9) and usage of mechanical ventilators (OR = 3.9; 95% CI: 1.3, 11.9) P < 0.02) (Hortiwakul et al., 2012).
Many studies were carried out to explore the influence of microbiological factors on the outcome of HA-BSI, and some factors were found to be associated with high mortality rates among HA-BSI cases, like a causative pathogen, poly-microbial infection, and the source of infection. According to our study, Stenotrophomonas maltophilia, Candida spp, Acinetobacter spp, and Pseudomonas spp were the leading pathogens associated with the highest pathogen-specific observed CFR among HA-BSI (43.5%, 41.1%. 40.8% and 40.5%, respectively. Infection by Stenotrophomonas maltophilia was associated with high rates of in-hospital deaths (49.7%) among patients with pneumonia (Guerci et al., 2019). The mortality rate among patients with HA-BSI due to Stenotrophomonas maltophilia was 60.6% (Chang et al., 2014). According to a Chinese study, the overall mortality rate among patients with Stenotrophomonas maltophilia BSI was 34.2% (Chen et al., 2019). A surveillance study from Saudia Arabia reported 50% mortality associated with Candida BSI (Al Thaqafi et al., 2014). Spanish study stated that 32% of BSI's mortality rate was due to Candida (Ortega et al., 2011). The rate obtained by our study lies in the middle between the two previous studies. Studies confirmed a significant association between Candida and Pseudomonas spp and high mortality rates among bacteremic patients ( Wu et al., 2015; Prowle et al., 2011; Kontula et al., 2018; Pittet et al., 1997). As per our findings, Corynebacterium spp, Enterobacter spp, Staphylococcus spp, and Serratia spp were more prevalent among survived HA-BSI patients causing pathogen-specific CFRs of less than 20%. In a study by Oliver, E.coli was associated with lower mortality rates compared to S. aureus (P < 0.0001) (Robineau et al., 2018). A New Zealander study found that mortality is lower among patients with HCA-BSI due to CoNS (6%) (Raymond et al., 2006). HA-BSI due to Gram-positive bacteria (58%) was frequently noted among patients who survived more than 28 days, specifically CoNS (23%) and S. aureus (16%), whereas E.coli (14%) and P. aerugenosa (12%) were the most frequent Gram-negative isolates causing HA-BSI resulting in early death (P < 0.0001). Fungal and polymicrobial BSIs were significantly associated with 28 days of mortality (Kontula et al., 2018). ICU-acquired BSIs due to Candida, S. aureus and Gram-negative bacilli were more frequently observed among dead cases (Prowle et al., 2011). In 9 public hospitals in Queensland, Australia, the risk of death HA-BSI is significantly increased when the infection is by methicillin-resistant Staphylococcus aureus (HR = 4.6, 95% CI; 2.7 to 7.6) (Barnett et al., 2013). Enterococcus faecium and Pseudomonas aeruginosa were associated with high mortality rates (Wu et al., 2015)