Performance of ICD-10-AM codes for quality improvement monitoring of hospital-acquired pneumonia in a haematology-oncology casemix in Victoria, Australia

BACKGROUND

The Australian hospital-acquired complication (HAC) policy was introduced to facilitate negative funding adjustments in Australian hospitals using ICD-10-AM codes.

OBJECTIVE

The aim of this study was to determine the positive predictive value (PPV) of the ICD-10-AM codes in the HAC framework to detect hospital-acquired pneumonia in patients with cancer and to describe any change in PPV before and after implementation of an electronic medical record (EMR) at our centre.

METHOD

A retrospective case review of all coded pneumonia episodes at the Peter MacCallum Cancer Centre in Melbourne, Australia spanning two time periods (01 July 2015 to 30 June 2017 [pre-EMR period] and 01 September 2020 to 28 February 2021 [EMR period]) was performed to determine the proportion of events satisfying standardised surveillance definitions.

RESULTS

HAC-coded pneumonia occurred in 3.66% (n = 151) of 41,260 separations during the study period. Of the 151 coded pneumonia separations, 27 satisfied consensus surveillance criteria, corresponding to an overall PPV of 0.18 (95% CI: 0.12, 0.25). The PPV was approximately three times higher following EMR implementation (0.34 [95% CI: 0.19, 0.53] versus 0.13 [95% CI: 0.08, 0.21]; p = .013).

CONCLUSION

The current HAC definition is a poor-to-moderate classifier for hospital-acquired pneumonia in patients with cancer and, therefore, may not accurately reflect hospital-level quality improvement. Implementation of an EMR did enhance case detection, and future refinements to administratively coded data in support of robust monitoring frameworks should focus on EMR systems.

IMPLICATIONS

Although ICD-10-AM data are readily available in Australian healthcare settings, these data are not sufficient for monitoring and reporting of hospital-acquired pneumonia in haematology-oncology patients.

Nursing and healthcare-associated pneumonia due to SARS-CoV-2 Omicron variant

BACKGROUND

There were many differences in the clinical characteristics between nursing and healthcare-associated pneumonia (NHCAP) and community-acquired pneumonia (CAP) due to the SARS-CoV-2 ancestral strain, Alpha variant and Delta variant. With the replacement of the Delta variant by the Omicron variant, the Omicron variant showed decreased infectivity to lung and was less pathogenic. We investigated the clinical differences between NHCAP and CAP due to the Omicron variant.

METHODS

We analyzed 516 NHCAP and 547 CAP patients with COVID-19 pneumonia. Of 516 patients with COVID-19 NHCAP, 330 cases were the Omicron variant (120 cases were BA.1, 53 cases were BA.2, and 157 cases were BA.5 subvariants) and 186 cases were non-Omicron variants.

RESULTS

The median age, frequency of comorbid illness, rates of intensive care unit (ICU) stay, and mortality rate were significantly higher in Omicron patients with NHCAP than in those with CAP. Rates of ICU stay and in-hospital mortality were significantly higher in NHCAP patients with non-Omicron variants compared with those in the Omicron variant group. No clinical differences were observed in patients with NHCAP among the Omicron BA.1, BA.2, and BA.5 subvariant groups.

CONCLUSIONS

The present study supported that the NHCAP category is necessary not only for bacterial pneumonia but also viral pneumonia. It is necessary to consider prevention and treatment strategies depending on the presence or absence of applicable criteria for NHCAP.

[Selected novelties for hospital based internal medicine]

In this selective overview of articles, we describe new concepts, therapeutic measures and pharmacological agents that may modify current practice in clinical internal medicine. Novelties for the management of cardiovascular disease, such as heart failure, hypoxemic respiratory failure, nosocomial pneumonia and certain allergies are discussed.

Delftia acidovorans sepsis in a neonate with nosocomial pneumonia

A term neonate weighing 2900 g was referred in the first week of life with complaints of abdominal distension since birth. New onset respiratory distress was noted 36 hours prior to referral. Baby required non-invasive respiratory support and intravenous antibiotics as chest X-ray was suggestive of pneumonia. Ultrasound abdomen confirmed low ano-rectal malformation treated with cut-back anoplasty. Blood culture on admission grew a rare organism Delftia acidovorans The antibiotics were tailored as per the sensitivity pattern and continued for a duration of 14 days. Respiratory distress gradually resolved by day 10 of admission. The baby developed pneumonia after 36 hours of stay in the referring hospital; most probably indicating a hospital acquired source of this pathogen. To the best of our knowledge, this is the second reported neonatal case of D. acidovorans sepsis.

Non-ventilator-associated ICU-acquired pneumonia (NV-ICU-AP) in patients with acute exacerbation of COPD: From the French OUTCOMEREA cohort

BACKGROUND

Non-ventilator-associated ICU-acquired pneumonia (NV-ICU-AP), a nosocomial pneumonia that is not related to invasive mechanical ventilation (IMV), has been less studied than ventilator-associated pneumonia, and never in the context of patients in an ICU for severe acute exacerbation of chronic obstructive pulmonary disease (AECOPD), a common cause of ICU admission. This study aimed to determine the factors associated with NV-ICU-AP occurrence and assess the association between NV-ICU-AP and the outcomes of these patients.

METHODS

Data were extracted from the French ICU database, OutcomeRea™. Using survival analyses with competing risk management, we sought the factors associated with the occurrence of NV-ICU-AP. Then we assessed the association between NV-ICU-AP and mortality, intubation rates, and length of stay in the ICU.

RESULTS

Of the 844 COPD exacerbations managed in ICUs without immediate IMV, NV-ICU-AP occurred in 42 patients (5%) with an incidence density of 10.8 per 1,000 patient-days. In multivariate analysis, prescription of antibiotics at ICU admission (sHR, 0.45 [0.23; 0.86], p = 0.02) and no decrease in consciousness (sHR, 0.35 [0.16; 0.76]; p < 0.01) were associated with a lower risk of NV-ICU-AP. After adjusting for confounders, NV-ICU-AP was associated with increased 28-day mortality (HR = 3.03 [1.36; 6.73]; p < 0.01), an increased risk of intubation (csHR, 5.00 [2.54; 9.85]; p < 0.01) and with a 10-day increase in ICU length of stay (p < 0.01).

CONCLUSION

We found that NV-ICU-AP incidence reached 10.8/1000 patient-days and was associated with increased risks of intubation, 28-day mortality, and longer stay for patients admitted with AECOPD.

Clinically defined aspiration pneumonia is an independent risk factor associated with long-term hospital stay: a prospective cohort study

BACKGROUND

Long-term hospital stay is associated with functional decline in patients with pneumonia, especially in the elderly. Among elderly patients with pneumonia, aspiration pneumonia is a major category. Clinical definition is usually used because it can occur without apparent aspiration episodes. It is still not clear whether a long-term hospital stay is due to aspiration pneumonia itself caused by underlying oropharyngeal dysfunction or simply due to functional decline in elderly patients with multiple comorbidities during acute infection. The aim of this study is to identify whether clinically defined aspiration pneumonia itself was associated with a long-term hospital stay.

METHODS

A prospective observational study on community-acquired (CAP) or healthcare-associated pneumonia (HCAP) was conducted from January 2012 through January 2014. Aspiration pneumonia was clinically defined as pneumonia not only occurring in patients after documented aspiration episodes, but also occurring in those with underlying oropharyngeal dysfunction: chronic disturbances of consciousness and/or chronic neuromuscular diseases. We defined thirty-day hospital stay as a long-term hospital stay and compared it with logistic regression analysis. Potential confounders included age, sex, HCAP, body mass index (BMI), long-term bed-ridden state, heart failure, cerebrovascular disorders, dementia, antipsychotics use, hypnotics use, and CURB score which is a clinical prediction tool used to assess the severity, standing for; C (presence of Confusion), U (high blood Urea nitrogen level), R (high Respiratory rate), and B (low Blood pressure). In a sub-analysis, we also explored factors associated with long-term hospital stay in patients with aspiration pneumonia.

RESULTS

Of 2,795 patients, 878 (31.4%) had aspiration pneumonia. After adjusting potential confounders, the aspiration pneumonia itself was significantly associated with long-term hospital stay (adjusted odds ratio 1.44; 95% confidence interval 1.09-1.89, p < 0.01), as were higher age, male sex, high CURB score, HCAP, low BMI, heart failure, cerebrovascular disease, and antipsychotics use. Sub-analysis revealed factors associated with long-term hospital stay in the aspiration pneumonia, which included male sex, and multi-lobar chest X-ray involvement.

CONCLUSIONS

Clinically defined aspiration pneumonia itself was independently associated with long-term hospital stay. This result could potentially lead to specific rehabilitation strategies for pneumonia patients with underlying oropharyngeal dysfunction.

Positive Rate and Utility of Blood Culture among Nursing and Healthcare-associated Pneumonia Inpatients: A Cross-sectional Study

Objective Although blood cultures to identify the presence of bacteremia are recommended for nursing- and healthcare-associated pneumonia (NHCAP), the incidence of true bacteremia and the relationship between true bacteremia and the outcome remain unclear. Physicians can therefore sometimes be confused regarding whether or not blood cultures should be obtained for NHCAP patients. This study assessed the incidence of true bacteremia and the relationship between true bacteremia and the outcome of NHCAP in a Japanese hospital setting. Methods We retrospectively analyzed NHCAP patients hospitalized between April 2016 and March 2021. The primary outcome was the incidence of true bacteremia in blood cultures. The incidence of true bacteremia was also examined according to quick Sequential Organ Failure Assessment (qSOFA) and A-DROP scores. In addition, we compared the incidence of true bacteremia between survivors and non-survivors. Results In total, 205 patients were included in this study. Blood cultures were obtained from 150 of the 205 patients (73.2%). Positive blood cultures were detected in 26 patients (17.3%), of which only 8 cases (5.3%; 95% confidence interval, 2.3-10.2%) were considered true bacteremia. Trend analyses for the incidence of true bacteremia according to qSOFA and A-DROP scores did not show any statistically significant results (p=0.49 for qSOFA; p=0.14 for A-DROP). The proportion of true bacteremia cases did not differ significantly between survivors and non-survivors. Conclusions The incidence of true bacteremia among NHCAP patients was very low. A strategy for determining indications for obtaining blood cultures from NHCAP patients needs to be established.

Rapid Point-of-Care PCR Testing of Drug-Resistant Strains on Endotracheal Aspirate Samples: A Repurposed Effective Tool in the Stepwise Approach of Healthcare-Acquired Pneumonia-A Pilot Study

Healthcare-associated pneumonia (HCAP) is a common nosocomial infection with high morbidity and mortality. Culture-based detection of the etiologic agent and drug susceptibility is time-consuming, potentially leading to the inadequate use of broad-spectrum empirical antibiotic regimens. The aim was to evaluate the diagnostic capabilities of rapid point-of-care multiplex polymerase chain reaction (PCR) assays from the endotracheal aspirate of critically ill patients with HCAP. A consecutive series of 29 intensive care unit (ICU) patients with HCAP and a control group of 28 patients undergoing elective surgical procedures were enrolled in the study. The results of the PCR assays were compared to the culture-based gold standard. The overall accuracy of the PCR assays was 95.12%, with a sensitivity of 92.31% and a specificity of 97.67%. The median time was 90 min for the rapid PCR tests (p < 0.001), while for the first preliminary results of the cultures, it was 48 h (46-72). The overall accuracy for rapid PCR testing in suggesting an adequate antibiotic adjustment was 82.98% (95% CI 69.19-92.35%), with a specificity of 90% (95% CI 55.50-99.75%), a positive predictive value of 96.77% (95% CI 83.30-99.92%), and a negative predictive value of 56.25 (95% CII 29.88-80.25%). This method of rapid point-of-care PCR could effectively guide antimicrobial stewardship in patients with healthcare-acquired pneumonia.

Respiratory Viruses in Nosocomial Pneumonia: An Evolving Paradigm

Nosocomial pneumonia (NP) represents a leading cause of morbidity and mortality in hospitalized patients. Historically, clinicians have considered hospital-acquired pneumonia (HAP) and ventilator-associated pneumonia (VAP), which comprise NP, to be essentially bacterial processes. As such, patients suspected of having either HAP or VAP are initially treated with broad-spectrum antibiotics, and few clinicians search for a possible culprit virus. Recent reports which build on earlier studies, however, indicate that viruses likely play an important role in NP. Studies employing viral diagnostics as part of the evaluation for NP indicate that common respiratory viruses can spread nosocomially and lead to HAP and VAP. Similarly, studies of the general epidemiology of respiratory viral infections, such as influenza, respiratory syncytial virus, adenovirus, and rhinovirus, confirm that these pathogens are important causes of NP, especially among immunosuppressed and pediatric patients. More importantly, these more contemporary analyses reveal that one cannot, based on clinical characteristics, distinguish a viral from a bacterial cause of NP. Additionally, viral HAP and VAP result in crude mortality rates that rival or exceed those reported in bacterial NP. Rigorous prospective, multicenter trials are needed to confirm the significance of respiratory viruses in NP, as are studies of novel therapeutics for these viral infections.

Respiratory metagenomics: route to routine service

PURPOSE OF REVIEW

The coronavirus disease 2019 pandemic demonstrated broad utility of pathogen sequencing with rapid methodological progress alongside global distribution of sequencing infrastructure. This review considers implications for now moving clinical metagenomics into routine service, with respiratory metagenomics as the exemplar use-case.

RECENT FINDINGS

Respiratory metagenomic workflows have completed proof-of-concept, providing organism identification and many genotypic antimicrobial resistance determinants from clinical samples in <6 h. This enables rapid escalation or de-escalation of empiric therapy for patient benefit and reducing selection of antimicrobial resistance, with genomic-typing available in the same time-frame. Attention is now focussed on demonstrating clinical, health-economic, accreditation, and regulatory requirements. More fundamentally, pathogen sequencing challenges the traditional culture-orientated time frame of microbiology laboratories, which through automation and centralisation risks becoming increasingly separated from the clinical setting. It presents an alternative future where infection experts are brought together around a single genetic output in an acute timeframe, aligning the microbiology target operating model with the wider human genomic and digital strategy.

SUMMARY

Pathogen sequencing is a transformational proposition for microbiology laboratories and their infectious diseases, infection control, and public health partners. Healthcare systems that link output from routine clinical metagenomic sequencing, with pandemic and antimicrobial resistance surveillance, will create valuable tools for protecting their population against future infectious diseases threats.

Comparison of clinical outcomes in critically ill COVID-19 patients on mechanical ventilation with nosocomial pneumonia between Alpha and Omicron variants

BACKGROUND

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) causes coronavirus disease 2019 (COVID-19), a pandemic that has resulted in millions of deaths worldwide. Critically ill COVID-19 patients who require intubation and develop nosocomial pneumonia, commonly caused by gram-negative bacilli, have a higher mortality rate than those without nosocomial pneumonia.

OBJECTIVES

The aim of this study is to compare the clinical characteristics and outcomes and associated risk factors of Alpha and Omicron SARS-CoV-2 variants in critically ill patients on mechanical ventilation (MV) with nosocomial pneumonia.

DESIGN

This is a retrospective single-center cohort study.

METHODS

This observational study was conducted at Taipei Veterans General Hospital, Taiwan from May 2021 to September 2022. Critically ill patients who had confirmed SARS-CoV-2 infection and intubated on a MV with bacterial pneumonia were enrolled. Demographic data, laboratory results, and treatment information were collected and analyzed. In addition, clinical outcomes among different SARS-CoV-2 variants were examined.

RESULTS

This study included 94 critically ill COVID-19 patients who required intubation and intensive care unit (ICU) admission. The Alpha group had a longer duration of SARS-CoV-2 viral shedding, MV days, and ICU stay, while the Omicron group had older age, more comorbidities, higher APACHE II scores, and higher in-hospital mortality (47.0% versus 25.0%, p = 0.047). However, independent risk factors for in-hospital mortality included malignancy, lower serum albumin levels, and lack of Remdesivir treatment, except for the SARS-CoV-2 variant.

CONCLUSION

Our study discovered a higher in-hospital mortality rate in severe COVID-19 patients with MV and secondary pneumonia infected with the Omicron variant compared to the Alpha variant; however, real independent risk factors for in-hospital mortality are malignancy, lower serum albumin level, and lack of Remdesivir treatment.

Factors predicting mortality among patients with COVID-19 associated hospital acquired pneumonia: insights from a tertiary care center

Hospital acquired pneumonia (HAP) is a severe and dangerous complication in patients admitted with COVID-19, causing significant morbidity and mortality globally. However, the early detection and subsequent management of high-risk cases may prevent disease progression and improve clinical outcomes. This study was undertaken in order to identify predictors of mortality in COVID-19 associated HAP. A retrospective study was performed on all patients who were admitted to a tertiary care center with COVID-19 associated HAP from July 2020 till November 2020. Data was collected on relevant demographic, clinical and laboratory parameters to determine their association with in-hospital mortality; 1574 files were reviewed, out of which 162 were included in the final study. The mean age of subjects was 59.4±13.8 and a majority were male (78.4%). There were 71 (48.3%) mortalities in the study sample. Klebsiella pneumoniae (31.5%) and Pseudomonas aeruginosa (30.2%) were the most common organisms overall. Clinically significant growth of Aspergillus sp. was observed in 41 (29.0%) of patients. On univariate analysis, several factors were found to be associated with mortality, including male gender (p=0.04), D-dimers >1.3 mg/L (p<0.001), ferritin >1000 µg/mL (p<0.001), LDH >500I.U/mL (p<0.001) and procalcitonin >2.0 µg/mL (p<0.001). On multivariate analysis, ferritin >1000ng/mL, initial site of care in Special Care Units or Intensive Care Units, developing respiratory failure and developing acute kidney injury were factors independently associated with mortality in our patient sample. These results indicate that serum ferritin levels may be a potentially useful biomarker in the management of COVID-19 associated HAP.

Clinical features of nursing and healthcare-associated pneumonia due to COVID-19

Introduction

The objective of this study was to clarify the clinical differences between nursing and healthcare-associated pneumonia (NHCAP) and community-acquired pneumonia (CAP) due to COVID-19. We also investigated the clinical characteristics to determine whether there is a difference between the variant and non-variant strain in patients with NHCAP due to COVID-19. In addition, we analyzed the clinical outcomes in NHCAP patients with mental disorders who were hospitalized in a medical institution for treatment of mental illness.

Methods

This study was conducted at five institutions and assessed a total of 836 patients with COVID-19 pneumonia (154 cases were classified as NHCAP and 335 had lineage B.1.1.7.).

Results

No differences in patient background, clinical findings, disease severity, or outcomes were observed in patients with NHCAP between the non-B.1.1.7 group and B.1.1.7 group. The median age, frequency of comorbid illness, rates of intensive care unit stay, and mortality rate were significantly higher in patients with NHCAP than in those with CAP. Among the patients with NHCAP, the mortality rate was highest at 37.5% in patients with recent cancer treatment, followed by elderly or disabled patients receiving nursing care (24.3%), residents of care facilities (23.0%), patients receiving dialysis (13.6%), and patients in mental hospitals (9.4%).

Conclusions

Our results demonstrated that there were many differences in the clinical characteristics between NHCAP patients and CAP patients due to COVID-19. It is necessary to consider the prevention and treatment content depending on the presence or absence of applicable criteria for NHCAP.

Implementing Oral Care as a Nursing Intervention to Reduce Hospital-Acquired Pneumonia Across the United States Department of Veterans Affairs Healthcare System

Hospital-acquired pneumonia is a preventable complication. The primary source of pneumonia among hospitalized and long-term care residents is aspiration of bacteria present in the oral biofilm. Reducing the bacterial burden in the mouth through consistent oral care is associated with a reduction in the incidence of hospital-acquired pneumonia. Following a significant reduction in pneumonia among non-ventilated patients in the research pilots, the Veterans Health Administration deployed the evidence-based, nurse-led oral care intervention called Hospital Acquired Pneumonia Prevention by Engaging Nurses as quality improvement nationwide. In this article, nursing informatics experts on the team describe the design and implementation of process and outcome measures of Hospital-Acquired Pneumonia Prevention by Engaging Nurses and outline lessons learned. The team used standardized terms and observations embedded within the EHR documentation templates to measure the oral care intervention in acute care areas. They also developed a tracking system for hospital-acquired pneumonia cases among non-ventilated patients. In addition to improving patient safety and care quality, Hospital-Acquired Pneumonia Prevention by Engaging Nurses links evidence-based practice with nursing informatics principles to generate numerous opportunities to measure the value of nursing at the point of care. This initiative was reported using SQUIRE 2.0: Standards for QUality Improvement Reporting Excellence.

Simple steps to prevent hospital-acquired pneumonia in non-intubated patients: a quality improvement project

BACKGROUND

Hospital-acquired pneumonia (HAP) affects approximately 1.5% of UK inpatients. As well as leading to significant morbidity and mortality, HAP increases burden on hospitals by lengthening hospital stay. At a district general hospital in Kent, a quality improvement project (QIP) was designed that introduced simple preventive measures that could be implemented by ward nurses and allied health professionals.

METHODS

Three audit cycles studying a total of 222 inpatients on elderly care wards were undertaken over a 6-month period to assess staff compliance at various stages of the project, with interventions between each cycle. Actions included raising bedheads to 30°, sitting patients out of bed for meals, discouraging use of drinking straws, and regular mouth care.

RESULTS

Overall, improvements were seen in three of the measures. Considering the percentage of patients, there was a 23% increase in patients with bedheads >30°, 21% increase in use of adult feeding cups rather than straws, and 26% rise in patients sitting out of bed for meals.

CONCLUSION

The main objective of this QIP was to show that these simple yet potentially life-saving interventions are easy to implement on a busy ward, and the results have shown this to be true.

Incidence and Predictors of Nonventilator Hospital-Acquired Pneumonia in a Community Hospital

BACKGROUND

Nonventilator hospital-acquired pneumonia (NV-HAP) is a common hospital-acquired condition that is amenable to basic nursing care interventions.

PURPOSE

The purpose of this study was to determine the incidence of NV-HAP in a California community hospital and to identify the patient and nursing care factors including missed nursing care associated with its development.

METHODS

A retrospective study identified possible NV-HAP cases with ICD-10 (International Classification of Diseases, Tenth Revision) codes and then validated cases using Centers for Disease Control and Prevention confirmatory criteria.

RESULTS

The incidence of NV-HAP in our hospital was 0.64 cases per 1000 patient-days. Patient factors most strongly associated with NV-HAP were age (each year of increased age was associated with a 4% increased likelihood of developing NV-HAP) (OR = 1.04-1.07) and the presence of underlying disease, which reduced odds of developing NV-HAP by 36% (OR = 0.36; 95% CI, 0.12-0.98). Head-of-bed elevation reduced by 26% the odds of developing NV-HAP (OR = 0.26; 95% CI, 0.07-0.08).

CONCLUSIONS

NV-HAP can be predicted and potentially prevented. Paradoxically, the presence of underlying disease was not positively associated with the development of NV-HAP in this study.

Non-ventilator health care-associated pneumonia (NV-HAP): NV-HAP Risk Factors

Several risk factors have a significant impact on the development of nonventilator health care-associated pneumonia (NV-HAP). This section discusses both the modifiable and nonmodifiable risk factors contributing to the incidence of NV-HAP. Lack of mobility, malnutrition, high blood glucose, inhibition of gastric acid and central nervous system depressants underscored in this section. Efforts to treat, modify, or alter these risk factors as opportunities to decrease the incidence of NV-HAP. In addition, we to address and assess nonmodifiable risk factors.

Non-ventilator health care-associated pneumonia (NV-HAP): Pediatrics

There is a paucity of data on the prevalence, risk factors or prevention of nonventilator health care-associated pneumonia (NV-HAP) in children. As with adults, viral, fungal and bacterial infections can lead to NV-HAP, but surveillance definitions and case ascertainment remain a challenge. This section will review the known epidemiology of NV-HAP in the pediatric population, risk factors, surveillance, and prevention practices. Prevention strategies specific to pediatrics, aimed at reducing risk of viral transmission to hospitalized children will be discussed.

Deep Learning from Incomplete Data: Detecting Imminent Risk of Hospital-acquired Pneumonia in ICU Patients

Hospital acquired pneumonia (HAP) is the second most common nosocomial infection in the ICU and costs an estimated $3.1 billion annually. The ability to predict HAP could improve patient outcomes and reduce costs. Traditional pneumonia risk prediction models rely on a small number of hand-chosen signs and symptoms and have been shown to poorly discriminate between low and high risk individuals. Consequently, we wanted to investigate whether modern data-driven techniques applied to respective pneumonia cohorts could provide more robust and discriminative prognostication of pneumonia risk. In this paper we present a deep learning system for predicting imminent pneumonia risk one or more days into the future using clinical observations documented in ICU notes for an at-risk population (n = 1, 467). We show how the system can be trained without direct supervision or feature engineering from sparse, noisy, and limited data to predict future pneumonia risk with 96% Sensitivity, 72% AUC, and 80% F1-measure, outperforming SVM approaches using the same features by 20% Accuracy (relative; 12% absolute).

[Population-based incidence of hospitalised pneumococcal pneumonia in adults with distinct baseline risk strata in Catalonia throughout 2015, the EPIVAC study.]

OBJECTIVE

Pneumococcal pneumonia is a major public health problem, especially in high risk population and older adults. This study assessed the epidemiology of pneumococcal pneumonia requiring hospitalisation among adults in Catalonia.

METHODS

This is a population-based cohort study, including all individual ≥50 years-old assigned to the Institut Catala de la Salut (Catalonia, Spain), who were prospectively followed from 01/01/2015 to 31/12/2015. The Catalonian information system for the development of research in primary care (SIDIAP) was used to establish baseline characteristics and risk-strata of cohort members at study start: low-risk (immunocompetent persons without risk conditions), intermediate-risk (immunocompetent persons with at-risk condition) and high-risk (immunocompromising conditions). All hospitalisations from pneumococcal pneumonia occurred among cohort members within 2015 were collected from CMBD discharge data of 64 reference Catalonian hospitals.

RESULTS

Among the 2.025.730 cohort members the global incidence of hospitalised pneumococcal pneumonia was 82.8 cases per 100,000 population-year (9.8 for bacteremic and 72.9 for non- bacteremic cases). Incidence substantially increased by age (34.9 in 50-64 years vs 88.7 in 65-79 years vs 231.5 in >80 years; p<0.001) and baseline-risk stratum (30.1, 119.1 and 240.7 in low-, intermediate- and high-risk stratum, respectively; p<0.001). Overall case-fatality was 4.9% (7.7% in bacteremic vs 4.5% in non- bacteremic; p=0.059). In multivariable models, high-risk stratum and oldest age were the strongest predictors for invasive and non-invasive cases, respectively.

CONCLUSIONS

The burden of pneumococcal pneumonia among adults is intermediate-low in Catalonia, but large incidences emerge among oldest persons and immunocompromised subjects.