External Validation of Prediction Models for Pneumonia in Primary Care Patients with Lower Respiratory Tract Infection: An Individual Patient Data Meta-Analysis

Diagnostic accuracy of point-of-care tests for acute respiratory infection: a systematic review of reviews

Background

Acute respiratory infections are a common reason for consultation with primary and emergency healthcare services. Identifying individuals with a bacterial infection is crucial to ensure appropriate treatment. However, it is also important to avoid overprescription of antibiotics, to prevent unnecessary side effects and antimicrobial resistance. We conducted a systematic review to summarise evidence on the diagnostic accuracy of symptoms, signs and point-of-care tests to diagnose bacterial respiratory tract infection in adults, and to diagnose two common respiratory viruses, influenza and respiratory syncytial virus.

Methods

The primary approach was an overview of existing systematic reviews. We conducted literature searches (22 May 2023) to identify systematic reviews of the diagnostic accuracy of point-of-care tests. Where multiple reviews were identified, we selected the most recent and comprehensive review, with the greatest overlap in scope with our review question. Methodological quality was assessed using the Risk of Bias in Systematic Reviews tool. Summary estimates of diagnostic accuracy (sensitivity, specificity or area under the curve) were extracted. Where no systematic review was identified, we searched for primary studies. We extracted sufficient data to construct a 2 × 2 table of diagnostic accuracy, to calculate sensitivity and specificity. Methodological quality was assessed using the Quality Assessment of Diagnostic Accuracy Studies version 2 tool. Where possible, meta-analyses were conducted. We used GRADE to assess the certainty of the evidence from existing reviews and new analyses.

Results

We identified 23 reviews which addressed our review question; 6 were selected as the most comprehensive and similar in scope to our review protocol. These systematic reviews considered the following tests for bacterial respiratory infection: individual symptoms and signs; combinations of symptoms and signs (in clinical prediction models); clinical prediction models incorporating C-reactive protein; and biological markers related to infection (including C-reactive protein, procalcitonin and others). We also identified systematic reviews that reported the accuracy of specific tests for influenza and respiratory syncytial virus. No reviews were found that assessed the diagnostic accuracy of white cell count for bacterial respiratory infection, or multiplex tests for influenza and respiratory syncytial virus. We therefore conducted searches for primary studies, and carried out meta-analyses for these index tests. Overall, we found that symptoms and signs have poor diagnostic accuracy for bacterial respiratory infection (sensitivity ranging from 9.6% to 89.1%; specificity ranging from 13.4% to 95%). Accuracy of biomarkers was slightly better, particularly when combinations of biomarkers were used (sensitivity 80-90%, specificity 82-93%). The sensitivity and specificity for influenza or respiratory syncytial virus varied considerably across the different types of tests. Tests involving nucleic acid amplification techniques (either single pathogen or multiplex tests) had the highest diagnostic accuracy for influenza (sensitivity 91-99.8%, specificity 96.8-99.4%).

Limitations

Most of the evidence was considered low or very low certainty when assessed with GRADE, due to imprecision in effect estimates, the potential for bias and the inclusion of participants outside the scope of this review (children, or people in hospital).

Future work

Currently evidence is insufficient to support routine use of point-of-care tests in primary and emergency care. Further work must establish whether the introduction of point-of-care tests adds value, or simply increases healthcare costs.

Funding

This article presents independent research funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme as award number NIHR159948.

PLUS-IS-LESS project: Procalcitonin and Lung UltraSonography-based antibiotherapy in patients with Lower rESpiratory tract infection in Swiss Emergency Departments: study protocol for a pragmatic stepped-wedge cluster-randomized trial

BACKGROUND

Lower respiratory tract infections (LRTIs) are among the most frequent infections and a significant contributor to inappropriate antibiotic prescription. Currently, no single diagnostic tool can reliably identify bacterial pneumonia. We thus evaluate a multimodal approach based on a clinical score, lung ultrasound (LUS), and the inflammatory biomarker, procalcitonin (PCT) to guide prescription of antibiotics. LUS outperforms chest X-ray in the identification of pneumonia, while PCT is known to be elevated in bacterial and/or severe infections. We propose a trial to test their synergistic potential in reducing antibiotic prescription while preserving patient safety in emergency departments (ED).

METHODS

The PLUS-IS-LESS study is a pragmatic, stepped-wedge cluster-randomized, clinical trial conducted in 10 Swiss EDs. It assesses the PLUS algorithm, which combines a clinical prediction score, LUS, PCT, and a clinical severity score to guide antibiotics among adults with LRTIs, compared with usual care. The co-primary endpoints are the proportion of patients prescribed antibiotics and the proportion of patients with clinical failure by day 28. Secondary endpoints include measurement of change in quality of life, length of hospital stay, antibiotic-related side effects, barriers and facilitators to the implementation of the algorithm, cost-effectiveness of the intervention, and identification of patterns of pneumonia in LUS using machine learning.

DISCUSSION

The PLUS algorithm aims to optimize prescription of antibiotics through improved diagnostic performance and maximization of physician adherence, while ensuring safety. It is based on previously validated tests and does therefore not expose participants to unforeseeable risks. Cluster randomization prevents cross-contamination between study groups, as physicians are not exposed to the intervention during or before the control period. The stepped-wedge implementation of the intervention allows effect calculation from both between- and within-cluster comparisons, which enhances statistical power and allows smaller sample size than a parallel cluster design. Moreover, it enables the training of all centers for the intervention, simplifying implementation if the results prove successful. The PLUS algorithm has the potential to improve the identification of LRTIs that would benefit from antibiotics. When scaled, the expected reduction in the proportion of antibiotics prescribed has the potential to not only decrease side effects and costs but also mitigate antibiotic resistance.

TRIAL REGISTRATION

This study was registered on July 19, 2022, on the ClinicalTrials.gov registry using reference number: NCT05463406.

TRIAL STATUS

Recruitment started on December 5, 2022, and will be completed on November 3, 2024. Current protocol version is version 3.0, dated April 3, 2023.

Pulse Oximetry as an Aid to Rule Out Pneumonia among Patients with a Lower Respiratory Tract Infection in Primary Care

Guidelines recommend chest X-rays (CXRs) to diagnose pneumonia and guide antibiotic treatment. This study aimed to identify clinical predictors of pneumonia that are visible on a chest X-ray (CXR+) which could support ruling out pneumonia and avoiding unnecessary CXRs, including oxygen saturation. A secondary analysis was performed in a clinical trial that included patients with suspected pneumonia in Swiss primary care. CXRs were reviewed by two radiologists. We evaluated the association between clinical signs (heart rate > 100/min, respiratory rate ≥ 24/min, temperature ≥ 37.8 °C, abnormal auscultation, and oxygen saturation < 95%) and CXR+ using multivariate analysis. We also calculated the diagnostic performance of the associated clinical signs combined in a clinical decision rule (CDR), as well as a CDR derived from a large meta-analysis (at least one of the following: heart rate > 100/min, respiratory rate ≥ 24/min, temperature ≥ 37.8 °C, or abnormal auscultation). Out of 469 patients from the initial trial, 107 had a CXR and were included in this study. Of these, 26 (24%) had a CXR+. We found that temperature and oxygen saturation were associated with CXR+. A CDR based on the presence of either temperature ≥ 37.8 °C and/or an oxygen saturation level < 95% had a sensitivity of 69% and a negative likelihood ratio (LR-) of 0.45. The CDR from the meta-analysis had a sensitivity of 92% and an LR- of 0.37. The addition of saturation < 95% to this CDR increased the sensitivity (96%) and decreased the LR- (0.21). In conclusion, this study suggests that pulse oximetry could be added to a simple CDR to decrease the probability of pneumonia to an acceptable level and avoid unnecessary CXRs.

Machine learning-assisted prediction of pneumonia based on non-invasive measures

Background

Pneumonia is an infection of the lungs that is characterized by high morbidity and mortality. The use of machine learning systems to detect respiratory diseases via non-invasive measures such as physical and laboratory parameters is gaining momentum and has been proposed to decrease diagnostic uncertainty associated with bacterial pneumonia. Herein, this study conducted several experiments using eight machine learning models to predict pneumonia based on biomarkers, laboratory parameters, and physical features.

Methods

We perform machine-learning analysis on 535 different patients, each with 45 features. Data normalization to rescale all real-valued features was performed. Since it is a binary problem, we categorized each patient into one class at a time. We designed three experiments to evaluate the models: (1) feature selection techniques to select appropriate features for the models, (2) experiments on the imbalanced original dataset, and (3) experiments on the SMOTE data. We then compared eight machine learning models to evaluate their effectiveness in predicting pneumonia

Results

Biomarkers such as C-reactive protein and procalcitonin demonstrated the most significant discriminating power. Ensemble machine learning models such as RF (accuracy = 92.0%, precision = 91.3%, recall = 96.0%, f1-Score = 93.6%) and XGBoost (accuracy = 90.8%, precision = 92.6%, recall = 92.3%, f1-score = 92.4%) achieved the highest performance accuracy on the original dataset with AUCs of 0.96 and 0.97, respectively. On the SMOTE dataset, RF and XGBoost achieved the highest prediction results with f1-scores of 92.0 and 91.2%, respectively. Also, AUC of 0.97 was achieved for both RF and XGBoost models.

Conclusions

Our models showed that in the diagnosis of pneumonia, individual clinical history, laboratory indicators, and symptoms do not have adequate discriminatory power. We can also conclude that the ensemble ML models performed better in this study.

Clinical sign and biomarker-based algorithm to identify bacterial pneumonia among outpatients with lower respiratory tract infection in Tanzania

BACKGROUND

Inappropriate antibiotics use in lower respiratory tract infections (LRTI) is a major contributor to resistance. We aimed to design an algorithm based on clinical signs and host biomarkers to identify bacterial community-acquired pneumonia (CAP) among patients with LRTI.

METHODS

Participants with LRTI were selected in a prospective cohort of febrile (≥ 38 °C) adults presenting to outpatient clinics in Dar es Salaam. Participants underwent chest X-ray, multiplex PCR for respiratory pathogens, and measurements of 13 biomarkers. We evaluated the predictive accuracy of clinical signs and biomarkers using logistic regression and classification and regression tree analysis.

RESULTS

Of 110 patients with LRTI, 17 had bacterial CAP. Procalcitonin (PCT), interleukin-6 (IL-6) and soluble triggering receptor expressed by myeloid cells-1 (sTREM-1) showed an excellent predictive accuracy to identify bacterial CAP (AUROC 0.88, 95%CI 0.78-0.98; 0.84, 0.72-0.99; 0.83, 0.74-0.92, respectively). Combining respiratory rate with PCT or IL-6 significantly improved the model compared to respiratory rate alone (p = 0.006, p = 0.033, respectively). An algorithm with respiratory rate (≥ 32/min) and PCT (≥ 0.25 μg/L) had 94% sensitivity and 82% specificity.

CONCLUSIONS

PCT, IL-6 and sTREM-1 had an excellent predictive accuracy in differentiating bacterial CAP from other LRTIs. An algorithm combining respiratory rate and PCT displayed even better performance in this sub-Sahara African setting.

Respiratory tract infections (RTIs) in primary care: narrative review of C reactive protein (CRP) point-of-care testing (POCT) and antibacterial use in patients who present with symptoms of RTI

Antimicrobial resistance (AMR) continues to be a global problem and continues to be addressed through national strategies to improve diagnostics, develop new antimicrobials and promote antimicrobial stewardship. Patients who attend general (ambulatory) practice with symptoms of respiratory tract infections (RTIs) are invariably assessed by some sort of clinical decision rule (CDR). However, CDRs rely on a cluster of non-specific clinical observations. A narrative review of the literature was undertaken to ascertain the value of C reactive protein (CRP) point-of-care testing (POCT) to guide antibacterial prescribing in adult patients presenting to general practitioner (GP) practices with symptoms of RTI. Studies that were included were Cochrane reviews, systematic reviews, randomised controlled trials, cluster randomised trials, controlled before and after studies, cohort studies and economic evaluations. An overwhelming number of studies demonstrated that the use of CRP tests in patients presenting with RTI symptoms reduces index antibacterial prescribing. GPs and patients report a good acceptability for a CRP POCT and economic evaluations show cost-effectiveness of CRP POCT over existing RTI management in primary care. POCTs increase diagnostic precision for GPs in the better management of patients with RTI. With the rapid development of artificial intelligence, patients will expect greater precision in diagnosing and managing their illnesses. Adopting systems that markedly reduce antibiotic consumption is a no-brainer for governments that are struggling to address the rise in AMR.

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).

Community-acquired pneumonia in the emergency department: an algorithm to facilitate diagnosis and guide chest CT scan indication

OBJECTIVE

The aim was to create and validate a community-acquired pneumonia (CAP) diagnostic algorithm to facilitate diagnosis and guide chest computed tomography (CT) scan indication in patients with CAP suspicion in Emergency Departments (ED).

METHODS

We performed an analysis of CAP suspected patients enrolled in the ESCAPED study who had undergone chest CT scan and detection of respiratory pathogens through nasopharyngeal PCRs. An adjudication committee assigned the final CAP probability (reference standard). Variables associated with confirmed CAP were used to create weighted CAP diagnostic scores. We estimated the score values for which CT scans helped correctly identify CAP, therefore creating a CAP diagnosis algorithm. Algorithms were externally validated in an independent cohort of 200 patients consecutively admitted in a Swiss hospital for CAP suspicion.

RESULTS

Among the 319 patients included, 51% (163/319) were classified as confirmed CAP and 49% (156/319) as excluded CAP. Cough (weight = 1), chest pain (1), fever (1), positive PCR (except for rhinovirus) (1), C-reactive protein ≥50 mg/L (2) and chest X-ray parenchymal infiltrate (2) were associated with CAP. Patients with a score below 3 had a low probability of CAP (17%, 14/84), whereas those above 5 had a high probability (88%, 51/58). The algorithm (score calculation + CT scan in patients with score between 3 and 5) showed sensitivity 73% (95% CI 66-80), specificity 89% (95% CI 83-94), positive predictive value (PPV) 88% (95% CI 81-93), negative predictive value (NPV) 76% (95% CI 69-82) and area under the curve (AUC) 0.81 (95% CI 0.77-0.85). The algorithm displayed similar performance in the validation cohort (sensitivity 88% (95% CI 81-92), specificity 72% (95% CI 60-81), PPV 86% (95% CI 79-91), NPV 75% (95% CI 63-84) and AUC 0.80 (95% CI 0.73-0.87).

CONCLUSION

Our CAP diagnostic algorithm may help reduce CAP misdiagnosis and optimize the use of chest CT scan.

Clinical features for diagnosis of pneumonia among adults in primary care setting: A systematic and meta-review

Pneumonia results in significant morbidity and mortality worldwide. However, chest radiography may not be accessible in primary care setting. We aimed to evaluate clinical features and its diagnostic value to identify pneumonia among adults in primary care settings. Three academic databases were searched and included studies that assessed clinical predictors of pneumonia, adults without serious illness, have CXR and have conducted in primary care settings. We calculated sensitivity, specificity, positive and negative likelihood ratios, diagnostic odds ratio of each index test and the pool estimates for index tests. We identified 2,397 articles, of which 13 articles were included. In our meta-analysis, clinical features with the best pooled positive likelihood ratios were respiratory rate ≥20 min⁻¹ (3.47; 1.46–7.23), temperature ≥38 °C (3.21; 2.36–4.23), pulse rate >100 min⁻¹ (2.79; 1.71–4.33), and crackles (2.42; 1.19–4.69). Laboratory testing showed highest pooled positive likelihood ratios with PCT >0.25 ng/ml (7.61; 3.28–15.1) and CRP > 20 mg/l (3.76; 2.3–5.91). Cough, pyrexia, tachycardia, tachypnea, and crackles are limited as a single predictor for diagnosis of radiographic pneumonia among adults. Development of clinical decision rule that combine these clinical features together with molecular biomarkers may further increase overall accuracy for diagnosis of radiographic pneumonia among adults in primary care setting.

Predictors of Adverse Outcomes in Uncomplicated Lower Respiratory Tract Infections

PURPOSE

Presentation with acute lower respiratory tract infection (LRTI) in primary care is common. The aim of this study was to help clinicians treat patients presenting with LRTI in primary care by identifying those at risk of serious adverse outcomes (death, admission, late-onset pneumonia).

METHODS

In a prospective cohort study of patients presenting with LRTI symptoms, patient characteristics and clinical findings were recorded and adverse events identified over 30 days by chart review. Multivariable logistic regression analyses identified predictors of adverse outcomes.

RESULTS

Participants were recruited from 522 UK practices in 2009-2013. The analysis was restricted to the 28,846 adult patients not referred immediately to the hospital. Serious adverse outcomes occurred in 325/28,846 (1.1%). Eight factors were independently predictive; these characterized symptom severity (absence of coryza, fever, chest pain, and clinician-assessed severity), patient vulnerability (age >65 years, comorbidity), and physiological impact (oxygen saturation <95%, low blood pressure). In aggregate, the 8 features had moderate predictive value (area under the receiver operating characteristic curve 0.71, 95% CI, 0.68-0.74); the 4% of patients with ≥5 features had an approximately 1 in 17 (5.7%) risk of serious adverse outcomes, the 35% with 3 or 4 features had an intermediate risk (1 in 50, 2.0%), and the 61% with ≤2 features had a low (1 in 200, 0.5%) risk.

CONCLUSIONS

In routine practice most patients presenting with LRTI in primary care can be identified as at intermediate or low risk of serious outcome.

Rational Use of CT-Scan for the Diagnosis of Pneumonia: Comparative Accuracy of Different Strategies

Diagnosing pneumonia in emergency departments is challenging because the accuracy of symptoms, signs and laboratory tests is limited. As a confirmation test, chest X-ray has significant limitations and is outperformed by CT-scan. However, obtaining a CT-scan in all cases of suspected pneumonia has significant drawbacks. We used a cohort of 200 consecutive elderly patients admitted to the hospital for suspected pneumonia to build a simple prediction score, which was used to determine indication for performing a CT-scan. The reference diagnosis was adjudicated by experts considering all available data, including evolution until discharge and CT scan in all patients. Results were externally validated in a second cohort of 319 patients. Pneumonia was confirmed in 133 patients (67%). Area under the receiver operator curve (AUROC) of physician evaluation was 0.55 (0.46-0.64). The score incorporated four variables independently predicting confirmed pneumonia: male gender, acute cough, C-reactive protein >70 mg/L, and urea <7 mmol/L. AUROC of the score was 0.68 (95% confidence interval (CI) 0.60-0.76). When a CT-scan was obtained for patients at low or intermediate predicted risk (108 patients, 54% of the cohort), AUROC was 0.71 (0.63-0.80) and 0.69 (0.64-0.74) in the derivation and validation cohort, respectively. A simple prediction score for pneumonia had moderate accuracy and could guide the performance of a CT-scan.

Clinical prediction rule to predict pneumonia in adult presented with acute febrile respiratory illness

OBJECTIVE

To derive a clinical prediction rule to predict pneumonia in patients with acute febrile respiratory illness to emergency departments.

METHOD

This was a prospective multicentre study. 537 adults were recruited. Those requiring resuscitation or were hypoxaemic on presentation were excluded. Pneumonia was defined as new onset infiltrates on chest X-ray (CXR), or re-attendance within 7 days and diagnosed clinically as having pneumonia. A predictive model, the Acute Febrile Respiratory Illness (AFRI) rule was derived by logistic regression analysis based on clinical parameters. The AFRI rule was internally validated with bootstrap resampling and was compared with the Diehr and Heckerling rule.

RESULTS

In the 363 patients who underwent CXR, 100 had CXR confirmed pneumonia. There were 7 weighted factors within the ARFI rule, which on summation, gave the AFRI score: age ≥ 65 (1 point), peak temperature within 24 h ≥ 40 °C (2 points), fever duration ≥3 days (2 points), sore throat (-2 points), abnormal breath sounds (1 point), history of pneumonia (1 point) and SpO2 ≤ 96% (1 point). With the bootstrap resampling, the AFRI rule was found to be more accurate than the Diehr and Heckerling rule (area under ROC curve 0.816, 0.721 and 0.566 respectively, p < 0.001). At a cut-off of AFRI≥0, the rule was found to have 95% sensitivity, with a negative predictive value of 97.2%. Using the AFRI score, we found CXR could be avoided for patients having a score of <0.

CONCLUSION

AFRI score could assist emergency physicians in identifying pneumonia patients among all adult patients presented to ED for acute febrile respiratory illness.

Predictors of pneumonia in lower respiratory tract infections: 3C prospective cough complication cohort study

The aim was to aid diagnosis of pneumonia in those presenting with lower respiratory tract symptoms in routine primary care.

A cohort of 28 883 adult patients with acute cough attributed to lower respiratory tract infections (LRTIs) was recruited from 5222 UK practices in 2009–13. Symptoms, signs and treatment were recorded at presentation and subsequent events followed-up for 30 days by chart review. The predictive value of patient characteristics, presenting symptoms and clinical findings for the diagnosis of pneumonia in the first 7 days was established.

Of the 720 out of 28 883 (2.5.%) radiographed within 1 week of the index consultation, 115 (16.0%; 0.40% of 28 883) were assigned a definite or probable pneumonia diagnosis. The significant independent predictors of radiograph-confirmed pneumonia were temperature >37.8°C (RR 2.6; 95% CI 1.5–4.8), crackles on auscultation (RR 1.8; 1.1–3.0), oxygen saturation <95% (RR 1.7; 1.0–3.1) and pulse >100·min^–1 (RR 1.9; 1.1–3.2). Most patients with pneumonia (99/115, 86.1%) exhibited at least one of these four clinical signs; the positive predictive value of having at least one of these signs was 20.2% (95% CI 17.3–23.1).

In routine practice, radiograph-confirmed pneumonia as a short-term complication of LRTI is very uncommon (one in 270). Pulse oximetry may aid the diagnosis of pneumonia in this setting.

Pulse oximetry probably has a role in the diagnosis of pneumonia in the communityhttp://ow.ly/QpWc30fVM2j

Diagnostic utility of appetite loss in addition to existing prediction models for community-acquired pneumonia in the elderly: a prospective diagnostic study in acute care hospitals in Japan

OBJECTIVE

Diagnosis of community-acquired pneumonia (CAP) in the elderly is often delayed because of atypical presentation and non-specific symptoms, such as appetite loss, falls and disturbance in consciousness. The aim of this study was to investigate the external validity of existing prediction models and the added value of the non-specific symptoms for the diagnosis of CAP in elderly patients.

DESIGN

Prospective cohort study.

SETTING

General medicine departments of three teaching hospitals in Japan.

PARTICIPANTS

A total of 109 elderly patients who consulted for upper respiratory symptoms between 1 October 2014 and 30 September 2016.

MAIN OUTCOME MEASURES

The reference standard for CAP was chest radiograph evaluated by two certified radiologists. The existing models were externally validated for diagnostic performance by calibration plot and discrimination. To evaluate the additional value of the non-specific symptoms to the existing prediction models, we developed an extended logistic regression model. Calibration, discrimination, category-free net reclassification improvement (NRI) and decision curve analysis (DCA) were investigated in the extended model.

RESULTS

Among the existing models, the model by van Vugt demonstrated the best performance, with an area under the curve of 0.75(95% CI 0.63 to 0.88); calibration plot showed good fit despite a significant Hosmer-Lemeshow test (p=0.017). Among the non-specific symptoms, appetite loss had positive likelihood ratio of 3.2 (2.0-5.3), negative likelihood ratio of 0.4 (0.2-0.7) and OR of 7.7 (3.0-19.7). Addition of appetite loss to the model by van Vugt led to improved calibration at p=0.48, NRI of 0.53 (p=0.019) and higher net benefit by DCA.

CONCLUSIONS

Information on appetite loss improved the performance of an existing model for the diagnosis of CAP in the elderly.

Prediction of pneumonia hospitalization in adults using health checkup data

Objectives

Community-acquired pneumonia is a common cause of hospitalization, and pneumococcal vaccinations are recommended for high-risk individuals. Although risk factors for pneumonia have been identified, there are currently no pneumonia hospitalization prediction models based on the risk profiles of healthy subjects. This study aimed to develop a predictive model for pneumonia hospitalization in adults to accurately identify high-risk individuals to facilitate the efficient prevention of pneumonia.

Methods

We conducted a retrospective database analysis using health checkup data and health insurance claims data for residents of Kyoto prefecture, Japan, between April 2010 and March 2015. We chose adults who had undergone health checkups in the first year of the study period, and tracked pneumonia hospitalizations over the next 5 years. Subjects were randomly divided into training and test sets. The outcome measure was pneumonia hospitalization, and candidate predictors were obtained from the health checkup data. The prediction model was developed and internally validated using a LASSO logistic regression analysis. Lastly, we compared the new model with comparative models.

Results

The study sample comprised 54,907 people who had undergone health checkups. Among these, 921 were hospitalized for pneumonia during the study period. The c-statistic for the prediction model in the test set was 0.71 (95% confidence interval: 0.69–0.73). In contrast, a comparative model with only age and comorbidities as predictors had a lower c-statistic of 0.55 (95% confidence interval: 0.54–0.56).

Conclusions

Our predictive model for pneumonia hospitalization performed better than comparative models, and may be useful for supporting the development of pneumonia prevention measures.