Evaluating the OpenAI's GPT-3.5 Turbo's performance in extracting information from scientific articles on diabetic retinopathy

We aimed to compare the concordance of information extracted and the time taken between a large language model (OpenAI's GPT-3.5 Turbo via API) against conventional human extraction methods in retrieving information from scientific articles on diabetic retinopathy (DR). The extraction was done using GPT3.5 Turbo as of October 2023. OpenAI's GPT-3.5 Turbo significantly reduced the time taken for extraction. Concordance was highest at 100% for the extraction of the country of study, 64.7% for significant risk factors of DR, 47.1% for exclusion and inclusion criteria, and lastly 41.2% for odds ratio (OR) and 95% confidence interval (CI). The concordance levels seemed to indicate the complexity associated with each prompt. This suggests that OpenAI's GPT-3.5 Turbo may be adopted to extract simple information that is easily located in the text, leaving more complex information to be extracted by the researcher. It is crucial to note that the foundation model is constantly improving significantly with new versions being released quickly. Subsequent work can focus on retrieval-augmented generation (RAG), embedding, chunking PDF into useful sections, and prompting to improve the accuracy of extraction.

Association between pre-biologic T2-biomarker combinations and response to biologics in patients with severe asthma

Background

To date, studies investigating the association between pre-biologic biomarker levels and post-biologic outcomes have been limited to single biomarkers and assessment of biologic efficacy from structured clinical trials.

Aim

To elucidate the associations of pre-biologic individual biomarker levels or their combinations with pre-to-post biologic changes in asthma outcomes in real-life.

Methods

This was a registry-based, cohort study using data from 23 countries, which shared data with the International Severe Asthma Registry (May 2017-February 2023). The investigated biomarkers (highest pre-biologic levels) were immunoglobulin E (IgE), blood eosinophil count (BEC) and fractional exhaled nitric oxide (FeNO). Pre- to approximately 12-month post-biologic change for each of three asthma outcome domains (i.e. exacerbation rate, symptom control and lung function), and the association of this change with pre-biologic biomarkers was investigated for individual and combined biomarkers.

Results

Overall, 3751 patients initiated biologics and were included in the analysis. No association was found between pre-biologic BEC and pre-to-post biologic change in exacerbation rate for any biologic class. However, higher pre-biologic BEC and FeNO were both associated with greater post-biologic improvement in FEV1 for both anti-IgE and anti-IL5/5R, with a trend for anti-IL4Rα. Mean FEV1 improved by 27-178 mL post-anti-IgE as pre-biologic BEC increased (250 to 1000 cells/µL), and by 43-216 mL and 129-250 mL post-anti-IL5/5R and -anti-IL4Rα, respectively along the same BEC gradient. Corresponding improvements along a FeNO gradient (25-100 ppb) were 41-274 mL, 69-207 mL and 148-224 mL for anti-IgE, anti-IL5/5R, and anti-IL4Rα, respectively. Higher baseline BEC was also associated with lower probability of uncontrolled asthma (OR 0.392; p=0.001) post-biologic for anti-IL5/5R. Pre-biologic IgE was a poor predictor of subsequent pre-to-post-biologic change for all outcomes assessed for all biologics. The combination of BEC + FeNO marginally improved the prediction of post-biologic FEV1 increase (adjusted R2: 0.751), compared to BEC (adjusted R2: 0.747) or FeNO alone (adjusted R2: 0.743) (p=0.005 and <0.001, respectively); however, this prediction was not improved by the addition of IgE.

Conclusions

The ability of higher baseline BEC, FeNO and their combination to predict biologic-associated lung function improvement may encourage earlier intervention in patients with impaired lung function or at risk of accelerated lung function decline.

Global health impacts of ambient fine particulate pollution associated with climate variability

Air pollution is a key global environmental problem raising human health concern. It is essential to comprehensively assess the long-term characteristics of air pollution and the resultant health impacts. We first assessed the global trends of fine particulate matter (PM2.5) during 1980-2020 using a monthly global PM2.5 reanalysis dataset, and evaluated their association with three types of climate variability including El Niño-Southern Oscillation, Indian Ocean Dipole and North Atlantic Oscillation. We then estimated PM2.5-attributable premature deaths using integrated exposure-response functions. Results show a significant increasing trend of ambient PM2.5 during 1980-2020 due to increases in anthropogenic emissions. Ambient PM2.5 caused a total of ∼ 135 million premature deaths globally during the four decades. Occurrence of air pollution episodes was strongly associated with climate variability, which were associated with up to 14 % increase in annual global PM2.5-attributable premature deaths.

A pilot randomised controlled trial exploring the feasibility and efficacy of a human-AI sleep coaching model for improving sleep among university students

Objective

Sleep quality is a crucial concern, particularly among youth. The integration of health coaching with question-answering (QA) systems presents the potential to foster behavioural changes and enhance health outcomes. This study proposes a novel human-AI sleep coaching model, combining health coaching by peers and a QA system, and assesses its feasibility and efficacy in improving university students' sleep quality.

Methods

In a four-week unblinded pilot randomised controlled trial, 59 university students (mean age: 21.9; 64% males) were randomly assigned to the intervention (health coaching and QA system; n = 30) or the control conditions (QA system; n = 29). Outcomes included efficacy of the intervention on sleep quality (Pittsburgh Sleep Quality Index; PSQI), objective and self-reported sleep measures (obtained from Fitbit and sleep diaries) and feasibility of the study procedures and the intervention.

Results

Analysis revealed no significant differences in sleep quality (PSQI) between intervention and control groups (adjusted mean difference = -0.51, 95% CI: [-1.55-0.77], p = 0.40). The intervention group demonstrated significant improvements in Fitbit measures of total sleep time (adjusted mean difference = 32.5, 95% CI: [5.9-59.1], p = 0.02) and time in bed (adjusted mean difference = 32.3, 95% CI: [2.7-61.9], p = 0.03) compared to the control group, although other sleep measures were insignificant. Adherence was high, with the majority of the intervention group attending all health coaching sessions. Most participants completed baseline and post-intervention self-report measures, all diary entries, and consistently wore Fitbits during sleep.

Conclusions

The proposed model showed improvements in specific sleep measures for university students and the feasibility of the study procedures and intervention. Future research may extend the intervention period to see substantive sleep quality improvements.

Density forecasting of conjunctivitis burden using high-dimensional environmental time series data

As one of the most common eye conditions being presented at clinics, acute conjunctivitis puts substantial strain on primary health resources. To reduce this public health burden, it is important to forecast and provide forward guidance to policymakers by estimating conjunctivitis trends, taking into account factors which influence transmission. Using a high-dimensional set of ambient air pollution and meteorological data, this study describes new approaches to point and probabilistic forecasting of conjunctivitis burden which can be readily translated to other infectious diseases. Over the period of 2012 - 2022, we show that simple models without environmental data provided better point forecasts but the more complex models which optimized predictive accuracy and combined multiple predictors demonstrated superior density forecast performance. These results were shown to be consistent over periods with and without structural breaks in transmission. Furthermore, ecological analysis using post-selection inference showed that increases in SO2, O3 surface concentration and total precipitation were associated to increased conjunctivitis attendance. The methods proposed can provide rich and informative forward guidance for outbreak preparedness and help guide healthcare resource planning in both stable periods of transmission and periods where structural breaks in data occur.

Forecasting upper respiratory tract infection burden using high-dimensional time series data and forecast combinations

Upper respiratory tract infections (URTIs) represent a large strain on primary health resources. To mitigate URTI transmission and public health burdens, it is important to pre-empt and provide forward guidance on URTI burden, while taking into account various facets which influence URTI transmission. This is so that appropriate public health measures can be taken to mitigate strain on primary care resources. This study describes a new approach to forecasting URTIs which can be used for national public health resource planning. Specifically, using environmental and disease data comprising more than 1000 dimensions, we developed sub-models which optimizes model explainability, in-sample model fit, predictive accuracy and combines many weaker predictors over a 2-month time horizon to generate direct, point forecasts over a 1-8 week ahead forecast horizon. Predictive performance was evaluated using rolling out-of-sample forecast assessment within both periods with/without structural breaks in transmission over the period of 2012-2022. We showed that forecast combinations of 5 other forecasting models had better and more consistent predictive performance than other modelling approaches, over periods with and without structural breaks in transmission dynamics. Furthermore, epidemiological analysis on high dimensional data was enabled using post-selection inference, to show the dynamic association between lower temperature, increases in past relative humidity and absolute humidity and increased URTIs attendance. The methods proposed can be used for outbreak preparedness and guide healthcare resource planning, in both stable periods of transmission and periods where structural breaks in data occur.

Impact of Palliative Care in End-of-Life of Fibrotic Interstitial Lung Disease Patients

Background: Interstitial lung disease (ILD) is associated with poor quality of life (QoL) and high symptom burden. Studies evaluating the benefits of palliative care examined mainly idiopathic pulmonary fibrosis (IPF) patients. We aim to examine the impact of palliative care on a broader group of fibrotic ILD patients. Methods: Single center retrospective cohort study comparing deceased ILD patients who received outpatient palliative care services (palliative-intervention group) against a usual care group. Results: Of 63 subjects, 26 (41%) were in the palliative-intervention group and 37 (59%) in the usual care group. Median time to palliative care referral was 8.6 (IQR .3-21.2) months. Dyspnea-related disability was greater in the palliative-intervention group [mMRC dyspnea score 3.5(IQR 2-4) vs 2(IQR 2-4), P = .039], with more patients requiring long term oxygen therapy (70% vs 30%, P < .001). There was no difference in the median number of hospitalizations or length of stay in the last 6 months of life. Patients in the palliative-intervention group had a higher uptake of advance care planning (ACP) (39% vs 11%, P = .014), lower frequency of intensive care unit (ICU) admissions (5% vs 19%, P = .102) and were prescribed more opioids (96% vs 27%, P < .001) and benzodiazepines (39% vs 14%, P = .022). The palliative-intervention group experienced a longer median survival of 23.9 months (95% confidence interval [CI] 14.1-33.7) compared to the usual group (11.4 months [95% CI 5.4-17.3] (log-rank test: P = .023). Male gender was a strong predictor of 1-year mortality. Conclusions: The palliative-intervention group received earlier pharmacologic intervention for symptom relief. Healthcare utilization was not increased despite greater dyspnea-related disability.

Restoration of HDAC2 and Nrf2 by andrographolide overcomes corticosteroid resistance in chronic obstructive pulmonary disease

BACKGROUND AND PURPOSE

Corticosteroid resistance poses a major barrier to an effective anti-inflammatory therapy for chronic obstructive pulmonary disease (COPD). The present study aimed to investigate potential corticosteroid re-sensitization actions of andrographolide, a bioactive molecule from the herb Andrographis paniculata, in COPD models, particularly in peripheral blood mononuclear cells (PBMCs) from COPD patients.

EXPERIMENTAL APPROACH

Corticosteroid sensitivity in PBMCs collected from COPD patients, or in human monocytic U937 cells exposed to cigarette smoke extract (CSE), was determined by measuring LPS-induced IL-8 production, in the presence and absence of andrographolide. The mechanisms of corticosteroid re-sensitization action of andrographolide were evaluated in a mouse cigarette smoke (CS)-induced acute lung injury model.

KEY RESULTS

Impaired inhibition of IL-8 production by dexamethasone was detected in PBMCs from COPD patients and in CSE-exposed U937 cells, together with reduced levels of nuclear factor erythroid 2-related factor 2 (Nrf2) and histone deacetylase-2 (HDAC2). In both PBMCs and CSE-exposed U937 cells, andrographolide restored dexamethasone inhibition of IL-8 production, accompanied by the up-regulation of Nrf2 and HDAC2 levels. In the U937 cells, andrographolide was able to block CSE-induced Akt and reduce the level of c-Jun. Besides, andrographolide also augmented dexamethasone actions on lowering total and neutrophil counts, cytokine levels, and oxidative damage markers in bronchoalveolar lavage fluid from CS-exposed mice.

CONCLUSION AND IMPLICATIONS

We report here for the first time a novel corticosteroid re-sensitization property of andrographolide in human PBMCs and provide mechanistic evidence to support clinical evaluation of andrographolide in reversing steroid resistance in COPD.

Long-term tolerance of a fractured self-expanding metal stent in a patient with adenoid cystic carcinoma

Self-expanding metals stent are used for recanalization of malignant CAO. Fracture of such a stent has been described. This prompts its removal. However removal of a metal stent is difficult and can be hazardous. Due to this US FDA advisory was released against their insertion in benign diseases where long life expectancy can expose the patients to the complications of prolonged indwelling time of these stents. We describe a case of adenoid cystic carcinoma who required SEMS placement for tracheal obstruction. However, stent fracture developed soon after its insertion causing patient to cough out stent pieces multiple times. In view of the challenge associated with removal of metal stent fixed to the tracheal walls at its proximal and distal ends, the stent was left in-situ. Patient has tolerated the fractured stent for 1-year and remains on follow-up without any complication or adverse event. This illustrates the long-term tolerance and safety of a fractured stent and helps to allay anxiety associated with this complication.

Outcome differences between recanalized malignant central airway obstruction from endoluminal disease versus extrinsic compression

We compared the outcome of Nd:YAG laser therapy with stent placement for malignant central airway obstruction (CAO) at our center over a 10-year period. This is a retrospective review of patients undergoing Nd:YAG laser therapy or self-expanding metal stent (SEMS) placement for malignant CAO between November 2007 and October 2017. Seventy-two patients were recanalized for malignant CAO. The median (range) age was 63 (23-86) years, with 49 (68%) males. Patients underwent either laser therapy alone (N = 36), stent placement alone (N = 30), or both (N = 6). The wavelength of Nd:YAG laser used was 1064 nm, and median (range) laser energy used was 25 (15-35) W, in 377 (115-1107) pulses. Fifty-one (71%) patients died with median survival of 7.2 months. In subgroup analysis, 21 (58.3%) vs. 25 (83.3%), p = 0.03 patients died in the "laser resection" vs. "stent placement" group with longer median survival of 12.4 months in the former vs. 4.5 months, p = 0.0004 in the later. Esophageal cancer and left main bronchus involvement were significantly more common (10 (33.3%) vs. 0, p = 0.0001, and 16 (53.3%) vs. 8 (22.2%), p = 0.01), in the stent placement vs. laser resection group, respectively. Trachea or main bronchi involvement and respiratory failure on presentation requiring mechanical ventilation correlated with poorer survival. The immediate restoration of luminal patency, complication rate, and 30-day mortality was similar among the two groups. The median (range) energy used for laser therapy was 25 (15-35) W. Median of 377 pulses was used for the duration of 287.5 s. The results were compared using a Wilcoxon two-sample test, and Fischer exact test with p values considered indicative of a significant difference if less than 0.05. In patients requiring recanalization of malignant CAO, the extrinsic compression from esophageal cancer, trachea or main bronchi involvement, respiratory failure on presentation requiring mechanical ventilation, and stent placement correlated with poorer survival. Interventional pulmonology training program should emphasize on dedicated training in laser therapy as it is associated with improved survival.

An Observational Study Evaluating the Performance of LENT Score in the Selected Population of Malignant Pleural Effusion from Lung Adenocarcinoma in Singapore

BACKGROUND

Patients with malignant pleural effusion (MPE) secondary to lung cancer have been associated with poor prognosis historically. LENT score developed to risk-stratify unselected patients with MPE predicts prognosis of < 6 months in patients with lung cancer.

OBJECTIVE

To assess the performance of LENT score in predicting prognosis in selected population of MPE secondary to lung adenocarcinoma alone.

METHODS

A retrospective observational study was conducted by reviewing the medical records of patients managed for MPE in the year 2012.

RESULTS

Seventy patients with lung adenocarcinoma presenting with MPE were studied. The median (range) LENT score at initial diagnosis was 5 (2-7), and the median survival 7.9 (0.13-40) months. Thirty-nine patients received epidermal growth factor receptor-tyrosine kinase inhibitor (EGFR-TKIs). The median LENT score and median survival was 4 (2-7) and 14.4 months, respectively, in this group. Those in high-risk category by LENT in this group (n = 19) had a median survival and 6-month survival of 17.4 months and 73.6%, respectively. Thirty-one patients were treated with conventional chemotherapy. The median LENT score and median survival was 5 (2-7) and 4.1 (0.13-34.3) months, respectively, in this group. The median survival and 6-month survival rate in patients in high-risk category and moderate-risk category by LENT score was 6.2 months and 52.7%, and 11.4 months and 70.5%, respectively.

CONCLUSION

LENT score underestimates prognosis in patients having MPE secondary to lung adenocarcinoma. This disparity particularly applies to the lung adenocarcinoma patients carrying EGFR mutation. Hence, LENT score may not be applicable to, or may need modification before applying to such patients.

Outcome of advanced lung cancer with central airway obstruction versus without central airway obstruction

Patients with central airway obstruction (CAO) from advanced lung cancer present with significant morbidity and are assumed to have lower survival. Hence, they are offered only palliative support. We asked if patients who have advanced lung cancer with CAO (recanalised and treated) will behave similarly to those with advanced lung cancer without CAO. This study was a retrospective review of the medical records of the patients managed for advanced lung cancer during 2010 and 2015 at our institution. 85 patients were studied. Median survival and 1-, 2- and 5-year survival were 5.8 months, 30.3%, 11.7% and 2.3% versus 9.3 months, 35.7%, 9.6% and 4.7%, respectively, in the CAO and no CAO groups (p=0.30). More patients presented with respiratory failure (15 (35%) versus none; p=0.0001) and required assisted mechanical ventilation (10 (23.3%) versus none; p=0.001) in the CAO group compared with the no CAO group. Fewer patients received chemotherapy in the CAO group (11 (25.5%)) compared with the no CAO group (23 (54.7%); p=0.008). There was no difference in survival among patients with advanced lung cancer whether they presented with CAO or without CAO. Survival was similar to those without CAO in patients with recanalised CAO despite greater morbidity and lesser use of chemotherapy, strongly advocating bronchoscopic recanalisation of CAO. These findings dispel the nihilism associated with such cases.

Can EGFR-Tyrosine Kinase Inhibitors (TKI) Alone Without Talc Pleurodesis Prevent Recurrence of Malignant Pleural Effusion (MPE) in Lung Adenocarcinoma

Abstract: Background and Objective: Epidermal Growth Factor Receptor-Tyrosine Kinase Inhibitors (EGFR-TKIs) are effective against lung adenocarcinoma. However, limited data is available assessing the effectiveness of EGFR-TKI use in preventing re-accumulation of MPE. To our knowledge, there is no literature on comparison of talc pleurodesis with EGFR-TKIs alone on re-accumulation of MPE in Asian population. We investigated if EGFR-TKI therapy for advanced lung adenocarcinoma with malignant pleural effusion (MPE) is also successful in preventing pleural fluid re-accumulation following initial drainage.

Methods: An observational cohort study of patients with lung adenocarcinoma and MPE in the year 2012 was conducted.

Results: 70 patients presented with MPE from lung adenocarcinoma. Fifty six underwent EGFR mutation testing of which 39 (69.6%) had activating EGFR mutation and 34 (87.1%) received TKI. 20 were managed by pleural fluid drainage only whereas 14 underwent talc pleurodesis following pleural fluid drainage. Time taken for the pleural effusion to re-accumulate in those with and without pleurodesis was 9.9 vs. 11.7 months, p=0.59 respectively. More patients (n=10, 25.6%) with activating EGFR mutation presented with complete opacification (white-out) of the hemithorax compared to none without activating EGFR mutation (p=0.02).

Conclusion: In TKI eligible patients, early talc pleurodesis may not confer additional benefit in preventing re-accumulation of pleural effusion and may be reserved for non-adenocarcinoma histology, or EGFR negative adenocarcinoma. Complete opacification of the hemithorax on presentation may serve as an early radiographic signal of positive EGFR mutation status.

Fentanyl-induced chest wall rigidity syndrome in a routine bronchoscopy

Combination of sedatives such as fentanyl and midazolam during bronchoscopy is recommended by American College of Chest Physician due to its favourable drug profile. It improves patient comfort and tolerance, and is commonly given unless contraindicated. We describe a rare case of fentanyl-induced chest wall rigidity syndrome during a routine bronchoscopy with endobronchial ultrasound guided-transbronchial needle aspiration (EBUS-TBNA) in a 55 year old male presenting with a lung mass and mediastinal lymphadenopathy. This was effectively managed with neuromuscular blockade, intubation and reversal agents including naloxone. This rare complication should be effectively managed by all bronchoscopist as it carries significant mortality and morbidity if not recognised early. We review the literature on the occurrence of fentanyl-induced chest wall rigidity and its predisposing risks factors.

Lung herniation after positive pressure ventilation

Lung herniation is a rare condition which has only been mainly published as case reports in the literature. Although its description has been around since 1845, not much is still known about it. More often seen after trauma or as a surgical complication post thoracic surgery, it's management usually depends on its etiology and varies between individuals. We report a case of lung herniation after positive pressure ventilation from a previous thoracotomy site. To our knowledge, this is the first report where the lung herniation was precipitated by mechanical ventilation.

Diagnostic performance of convex probe EBUS-TBNA in patients with mediastinal and coexistent endobronchial or peripheral lesions

To compare the performance of convex probe endobronchial ultrasound guided transbronchial needle aspiration (EBUS-TBNA) with conventional endobronchial biopsy (EBB) or transbronchial lung biopsy (TBLB) in patients with mediastinal, and coexisting endobronchial or peripheral lesions.Retrospective review of records of patients undergoing diagnostic EBUS-TBNA and conventional bronchoscopy in 2014.A total of 74 patients had mediastinal, and coexisting endobronchial or peripheral lesions. The detection rate of EBUS-TBNA for mediastinal lesion >1 cm in short axis, EBB for visible exophytic type of endobronchial lesion, and TBLB for peripheral lesion with bronchus sign were 71%, 75%, and 86%, respectively. In contrast, the detection rate of EBUS-TBNA for mediastinal lesion ≤1 cm in short axis, EBB for mucosal hyperemia type of endobronchial lesion, and TBLB for peripheral lesion without bronchus sign were 25%, 63%, and 38%, and improved to 63%, 88%, and 62% respectively by adding EBB or TBLB to EBUS-TBNA, and EBUS-TBNA to EBB or TBLB. Postprocedure bleeding was significantly more common in patients undergoing EBB and TBLB 8 (40%) versus convex probe EBUS-TBNA 2 patients (2.7%, P = 0.0004).EBUS-TBNA is a safer single diagnostic technique compared with EBB or TBLB in patients with mediastinal lesion of >1 cm in size, and coexisting exophytic type of endobronchial lesion, or peripheral lesion with bronchus sign. However, it requires combining with EBB or TBLB and vice versa to optimize yield when mediastinal lesion is ≤1 cm in size, and coexisting endobronchial and peripheral lesions lack exophytic nature, and bronchus sign, respectively.

Our Clinical Experience of Self-Expanding Metal Stent for Malignant Central Airway Obstruction

Background

We studied the safety, effectiveness, and limitations of airway stenting using self-expanding metal stent (SEMS) in patients with malignant central airway obstruction (CAO).

Methods

A retrospective review of records of patients undergoing SEMS placement for malignant CAO during year 2013 - 2014 was done.

Results

Sixteen patients (11 males and five females) underwent SEMS placement for malignant CAO. Median (range) age was 66 (54 - 78) years. No perioperative or immediate postoperative complications were seen except acute myocardial infarction (AMI) in one patient. Three patients were transferred to intensive care unit (ICU) for closer monitoring after the procedure and were discharged the next day. All four patients with lung atelectasis on presentation experienced complete re-expansion of the lung post-stenting. The dyspnea was substantially relieved in 14 (87.5%) patients. Two of the three patients who had been intubated were weaned off from the ventilator following stent insertion. Stent patency was maintained until death in all patients except one. Median survival from the date of diagnosis and the date of stent placement in lung cancer, esophageal cancer, and thyroid cancer were 140 (21 - 564) and 85 (15 - 361), 288 (80 - 419) and 61 (60 - 171), and 129 (71 - 187) and 67 (16 - 118) days, respectively. This survival was similar to reported expected survival associated with the underlying malignancy. During follow-up, granulation tissue (n = 1), mucostasis (n = 1), and tumor ingrowth (n = 2) were the most frequently encountered complications.

Conclusion

SEMSs are safe and effective in reversing respiratory failure caused by malignant CAO, averting premature death, allowing application of cancer targeted therapy, and restoring impending shortened survival to expected life expectancy associated with the underlying malignancy.

Erratum to: 36th International Symposium on Intensive Care and Emergency Medicine: Brussels, Belgium. 15-18 March 2016

[This corrects the article DOI: 10.1186/s13054-016-1208-6.].

Significance of coexistent granulomatous inflammation and lung cancer

Aims

Coexistence of lung cancer and granulomatous inflammation in the same patient confuses clinicians. We aimed to document the prevalence, clinicopathological features, treatment outcomes and prognosis in patients with coexisting granulomatous inflammation undergoing curative lung resection for lung cancer, in a tuberculosis (TB)-endemic country.

Methods

An observational cohort study of patients with lung cancer undergoing curative resection between 2012 and 2015 in a tertiary centre in Singapore.

Results

One hundred and twenty-seven patients underwent lung resection for cancer, out of which 19 (14.9%) had coexistent granulomatous inflammation in the resected specimen. Median age was 68 years and 58.2% were males. Overall median (range) survival was 451 (22–2452) days. Eighteen (14%) patients died at median duration of 271 days after surgery. The postsurgery median survival for those alive was 494 (29–2452) days in the whole group. Subgroup analysis did not reveal any differences in age, gender, location of cancer, radiological features, type of cancer, chemotherapy, history of TB or survival in patients with or without coexistent granulomatous inflammation.

Conclusions

Incidental detection of granulomatous inflammation in patients undergoing lung resection for cancer, even in a TB-endemic country, may not require any intervention. Such findings may be due to either mycobacterial infection in the past or ‘sarcoid reaction’ to cancer. Although all patients should have their resected specimen sent for acid-fast bacilli culture and followed up until the culture results are reported, the initiation of the management of such patients as per existing lung cancer management guidelines does not affect their outcome adversely.

Evaluation of a disease management program for COPD using propensity matched control group

BACKGROUND

Disease management programs (DMPs) have proliferated recently as a means of improving the quality and efficiency of care for patients with chronic illness. These programs include education about disease, optimization of evidence-based medications, information and support from case managers, and institution of self-management principles. Chronic obstructive pulmonary disease (COPD) is a major cause of morbidity and mortality in Singapore and worldwide. DMP aims to reduce mortality, hospitalizations, and average length of stay in such patients. This study assesses the outcomes of the DMP, comparing the propensity score matched DMP patients with controls.

METHODS

DMP patients were compared with the controls, who were COPD patients fulfilling the DMP's inclusion criteria but not included in the program. Control patients were identified from Operations Data Store (ODS) database. The outcomes of interest were average length of stay, number of days admitted to hospital per 100 person days, readmission, and mortality rates per person year. The risk of death and readmission was estimated using Cox, and competing risk regression respectively. Propensity score was estimated to identify the predictors of DMP enrolment. DMP patients and controls were matched on their propensity score.

RESULTS

There were 170 matched DMP patients and control patients having 287 and 207 hospitalizations respectively. Program patient had lower mortality than the controls (0.12 vs. 0.27 per person year); cumulative 1-year survival was 91% among program patient and 76% among the control patients. Readmission, and hospital days per 100 person-days was higher for the program patients (0.36 vs. 0.17 per person year), and (2.19 vs. 1.88 per person year) respectively.

CONCLUSIONS

Participation in "DMP" was associated with lower all-cause mortality when compared to the controls. This survival gain in the program patients was paradoxically associated with an increase in readmission rate and total hospital days.

Pleural LDH as a prognostic marker in adenocarcinoma lung with malignant pleural effusion

To study the performance of serum and pleural lactate dehydrogenase (LDH) level in predicting survival in patients with adenocarcinoma lung presenting with malignant pleural effusions (MPE) at initial diagnosis.Retrospective cohort study of the patient hospitalized for adenocarcinoma lung with MPE in year 2012.Univariate analyses showed lower pleural fluid LDH 667 (313-967) versus 971 (214-3800), P = 0.04, female gender 9 (100%) versus 27 (41.5%), P = 0.009, never smoking status 9 (100%) versus 36 (55.3%), P = 0.009, and epidermal growth factor receptor-tyrosine kinase inhibitor (EGFR-TKI) therapy 8 (89%) versus 26 (40%), P = 0.009 to correlate with survival of more than 1.7 year versus less than 1.7 year. In multivariate analysis, low pleural fluid LDH and female gender maintained significance. The pleural LDH level of ≤1500 and >1500 U/L discriminated significantly (P = 0.009) between survival.High pleural LDH (>1500 IU/L) predicts shorter survival (less than a year) in patients with adenocarcinoma lung presenting with MPE at the time of initial diagnosis. This marker may be clinically applied for selecting therapeutic modality directed at prevention of reaccumulation of MPE. Patients with low pleural LDH may be considered suitable for measures that provide more sustained effect on prevention of reaccumulation such as chemical pleurodesis or tunneled pleural catheter.

Evaluation of pulmonary nodules in Asian population

BACKGROUND

American College of Chest Physicians (ACCP) provides guidelines to manage pulmonary nodules. Pulmonary nodules however can be malignant or benign. Similar incidence of tuberculosis (TB) and lung cancer in Asian countries raises concern over the relevance of suggested guidelines in Asian population. There is little data on the pattern of clinical practice in the management of pulmonary nodules in Asian country (Singapore). Our study describes the current pattern of clinical practice in this area highlighting the variation in practice and discussing the potential reasons.

METHODS

Retrospective review of the medical records of patients diagnosed with lung cancer in 2010.

RESULTS

Sixty nodules were identified in 32 patients. Nodules were detected incidentally on routine imaging in 7 (21.9%) patients. TB contact tracing and pre-employment screening were common ways by which nodules were detected incidentally. Over one third (37.5%) were non-smokers. Majority of nodules were located in the upper lobes of right and left lung followed by right lower lobe (RLL). Only few patients 8 (25%) had positron emission tomography (PET) scan for staging purposes. There were no difference in survival between patients who presented with single, 747 (range, 25-1,840) days vs. multiple nodules 928 (range, 30-2,572) days, P=0.26. In a retrospective analysis of malignancy risk with the probability calculator, 62.5% patients were at low-moderate risk whilst 32.5% were at high risk.

CONCLUSIONS

The clinical practice of managing pulmonary nodules in Asian population differs from ACCP guidelines. None of the patient had pre-test probability calculated, and few had PET scan. This is because upper lobe predominance of lung cancer is identical to TB, non-smoking history does not have any weight in discounting malignancy risk where many of the Asian lung cancer patients are non-smokers, and the local endemicity of TB and its confounding effect on radiological findings of CT scan and PET scan.

Subconjunctival haemorrhage from bronchoscopy: A case report

Flexible bronchoscopy has been available for almost five decades. It has evolved as one of the most commonly used invasive diagnostic and therapeutic procedure in pulmonology, and its scope of applications is progressively expanding with the addition of new adjunct technologies such as endobronchial ultrasound, bronchial Thermoplasty, and navigational bronchoscopy. It is a safe procedure with complications ranging from fever, infiltrates, hypoxemia, bleeding, pneumothoraces and death, with most significant complications being bleeding and pneumothorax. We report a case of subconjuctival haemorrhage as an immediate complication of bronchoscopy. To our knowledge this is the first report documenting this rare complication.

Timeliness of Diagnosing Lung Cancer: Number of Procedures and Time Needed to Establish Diagnosis: Being Right the First Time

To study number of procedures and time to diagnose lung cancer and factors affecting the timeliness of clinching this diagnosis. Retrospective cohort study of lung cancer patients who consecutively underwent diagnostic bronchoscopy in 1 year (October 2013 to September 2014). Out of 101 patients diagnosed with lung cancer from bronchoscopy, average time interval between first abnormal computed tomogram (CT) scan-to-1st procedure, 1st procedure-to-diagnosis, and 1st abnormal CT scan-to-diagnosis was 16 ± 26, 11 ± 19, and 27 ± 33 days, respectively. These intervals were significantly longer in those requiring repeat procedures. Multivariate analysis revealed inconclusive 1st procedure to be the predictor of prolonged (>30 days) CT scan to diagnosis time (P = 0.04). Twenty-nine patients (28.7%) required repeat procedures (n = 63). Reasons behind repeating the procedures were inadequate procedure (n = 14), inaccessibility of lesion (n = 9), inappropriate procedure (n = 5), mutation analysis (n = 2), and others (n = 2). Fifty had visible endo-bronchial lesion, 20 had positive bronchus sign, and 83 had enlarged mediastinal/hilar lymph-nodes or central masses adjacent to the airways. Fewer procedures, and shorter procedure to diagnosis time, were observed in those undergoing convex probe endobronchial ultrasound-transbronchial needle aspiration (EBUS-TBNA) (P = 0.04). Most patients exhibit enlarged mediastinal lymph node or mass adjacent to the central airway accessible by convex probe EBUS-TBNA. Hence, combining it with conventional bronchoscopic techniques such as bronchoalveolar lavage, brush, and forceps biopsy increases detection rate, and reduces number of procedures and time to establish diagnosis. This may translate into cost and resource savings, timeliness of diagnosis, greater patient satisfaction, and conceivably better outcomes.

Predicting mortality among older adults hospitalized for community-acquired pneumonia: an enhanced confusion, urea, respiratory rate and blood pressure score compared with pneumonia severity index

BACKGROUND AND OBJECTIVE

Pneumonia Severity Index (PSI) predicts mortality better than Confusion, Urea >7 mmol/L, Respiratory rate >30/min, low Blood pressure: diastolic blood pressure <60 mm Hg or systolic blood pressure <90 mm Hg, and age >65 years (CURB-65) for community-acquired pneumonia (CAP) but is more cumbersome. The objective was to determine whether CURB enhanced with a small number of additional variables can predict mortality with at least the same accuracy as PSI.

METHODS

Retrospective review of medical records and administrative data of adults aged 55 years or older hospitalized for CAP over 1 year from three hospitals.

RESULTS

For 1052 hospital admissions of unique patients, 30-day mortality was 17.2%. PSI class and CURB-65 predicted 30-day mortality with area under curve (AUC) of 0.77 (95% confidence interval (CI): 0.73-0.80) and 0.70 (95% CI: 0.66-0.74) respectively. When age and three co-morbid conditions (metastatic cancer, solid tumours without metastases and stroke) were added to CURB, the AUC improved to 0.80 (95% CI: 0.77-0.83). Bootstrap validation obtained an AUC estimate of 0.78, indicating negligible overfitting of the model. Based on this model, a clinical score (enhanced CURB score) was developed that had possible values from 5 to 25. Its AUC was 0.79 (95% CI: 0.76-0.83) and remained similar to that of PSI class.

CONCLUSIONS

An enhanced CURB score predicted 30-day mortality with at least the same accuracy as PSI class did among older adults hospitalized for CAP. External validation of this score in other populations is the next step to determine whether it can be used more widely.

Economic burden of chronic obstructive pulmonary disease

BACKGROUND AND OBJECTIVE

The aim of this study was to estimate the direct medical costs of COPD in two public health clusters in Singapore from 2005 to 2009.

METHODS

Patients aged 40 years and over, who had been diagnosed with COPD, were identified in a Chronic Disease Management Data-mart. Annual utilization of health services in inpatient, specialist outpatient, emergency department and primary care settings was extracted from the Chronic Disease Management Data-mart. Trends in attributable costs, proportions of costs and health-care utilization were analyzed across each level of care. A weighted attribution approach was used to allocate costs to each health-care utilization episode, depending on the relevance of co-morbidities.

RESULTS

The mean total cost was approximately $9.9 million per year. Inpatient admissions were the major cost driver, contributing an average of $7.2 million per year. The proportion of hospitalization costs declined from 75% in 2005 to 68% in 2009. Based on the 5-year average, attendances at primary care clinics, emergency department and specialist clinics contributed 3%, 5% and 17%, respectively, of overall COPD costs. On average, 42% of the total cost burden was incurred for the medical management of COPD. The share of cost incurred for the treatment of conditions related and unrelated to COPD were 29% and 26%, respectively, of the total average costs.

CONCLUSIONS

COPD is likely to represent a significant burden to the public health system in most countries. The findings are particularly relevant to understanding the allocation of health-care resources and informing appropriate cost containment strategies.

Short-term mortality among older persons hospitalized for pneumonia: influence of baseline patient characteristics beyond severity of illness

BACKGROUND

Although severity of illness indices such as pneumonia severity index (PSI) are good predictors of short-term mortality for community-acquired pneumonia (CAP) and healthcare-associated pneumonia (HCAP), other patient factors may have added prognostic value.

OBJECTIVE

To identify patient factors beyond the PSI which explain 30-day mortality among older persons hospitalized with CAP and HCAP.

DESIGN

Retrospective cohort study.

SETTING

Three acute care hospitals in Singapore in 2007.

PATIENTS

Hospitalized adults aged 65 years or older who have primary International Classification of Diseases, 9th revision, Clinical Modification (ICD-9-CM) codes 480 to 486 with clinical and radiological features of pneumonia.

INTERVENTIONS

None.

MEASUREMENTS

Thirty-day mortality, PSI class, demographic and clinical features, comorbid conditions, functional status, selected laboratory tests, and chest radiographic findings.

RESULTS

Among 1607 patients included, 890 (55.4%) had CAP and 717 (44.6%) had HCAP. After adjustment for PSI class in logistic regression analyses, pre-morbid ambulation impairment (odds ratio [OR] 2.61, 95% confidence interval [CI] 1.98 to 3.45), hospitalization in the prior 30 days (OR 1.93, 95% CI 1.38 to 2.71), and absence of cough and purulent sputum (OR 1.47, 95% CI 1.14 to 1.90) were all significantly associated with 30-day mortality. These associations remained constant when CAP and HCAP were analyzed separately.

CONCLUSIONS

Recent hospitalization, pre-morbid ambulation impairment, and atypical presentation were independently associated with higher 30-day mortality among older persons hospitalized for pneumonia, after adjusting for severity of illness. These factors could be considered in addition to PSI when performing risk stratification and adjustment in this setting.

Difference in Asthma Control Test™ (ACT) Scores in Three Different Clinical Practice Settings

Introduction: Asthma control varies in different clinical settings because of its multidimensional and heterogeneous nature, and variability over time. The revised asthma management guidelines indicate that the goal of treatment should be maintaining asthma control for long periods. The aims of this study were to explore: (i) difference in asthma control test scores in patients at different clinical practice settings; (ii) assess if patients were overestimating the level of their asthma control and (iii) assess the relationship of the derived Asthma Control Test (ACT) score to cost of inpatient stay and length of stay (LOS). Materials and Methods: The Asthma Control Test (ACT) is a 5-item questionnaire that assesses the multidimensional perspective of asthma control from activity limitation, shortness of breath, night symptoms, use of rescue medication and self perception of asthma control. The score ranges on a scale from 1 (poorly controlled) to 5 (well controlled). ACT was administered to 447 patients diagnosed with asthma from the in-patient and out-patient settings (new and follow-up cases). Results: Three hundred and ninety-nine (92%) patients completed the ACT questionnaire. The analysis only included patients who had completed the ACT questionnaire. The analysis showed that all the 5 items in the ACT questionnaire were significantly associated with different clinical settings (P <0.001). When we correlated the ACT question 5 (patients self rating of asthma control) in the ACT with Question 3 and Question 4 individually, it showed that most patients did not overestimate their asthma control (P <0.001). However, there was no correlation between the derived ACT score and cost (P = 0.419), LOS (P = 0.373), and the number of comorbid medical history (P = 0.055). Conclusion: Our results reinforce the usefulness of ACT for clinicians to identify patients with poorly controlled asthma and to optimise their level of control in different clinical settings. Key words: Clinical guidelines, Overestimate, Useful

Difference in Asthma Control Test™ (ACT) scores in three different clinical practice settings

INTRODUCTION

Asthma control varies in different clinical settings because of its multidimensional and heterogeneous nature, and variability over time. The revised asthma management guidelines indicate that the goal of treatment should be maintaining asthma control for long periods. The aims of this study were to explore: (i) difference in asthma control test scores in patients at different clinical practice settings; (ii) assess if patients were overestimating the level of their asthma control and (iii) assess the relationship of the derived Asthma Control Test (ACT) score to cost of inpatient stay and length of stay (LOS).

MATERIALS AND METHODS

The Asthma Control Test (ACT) is a 5-item questionnaire that assesses the multidimensional perspective of asthma control from activity limitation, shortness of breath, night symptoms, use of rescue medication and self-perception of asthma control. The score ranges on a scale from 1 (poorly controlled) to 5 (well controlled). ACT was administered to 447 patients diagnosed with asthma from the in-patient and out-patient settings (new and follow-up cases).

RESULTS

Three hundred and ninety-nine (92%) patients completed the ACT questionnaire. The analysis only included patients who had completed the ACT questionnaire. The analysis showed that all the 5 items in the ACT questionnaire were significantly associated with different clinical settings (P <0.001). When we correlated the ACT question 5 (patients' self-rating of asthma control) in the ACT with Question 3 and Question 4 individually, it showed that most patients did not overestimate their asthma control (P <0.001). However, there was no correlation between the derived ACT score and cost (P = 0.419), LOS (P = 0.373), and the number of comorbid medical history (P = 0.055).

CONCLUSION

Our results reinforce the usefulness of ACT for clinicians to identify patients with poorly controlled asthma and to optimise their level of control in different clinical settings.

The use of text messaging to improve asthma control: a pilot study using the mobile phone short messaging service (SMS)

We have evaluated the feasibility of using the mobile phone short message service (SMS) for symptom monitoring in patients with asthma. All consecutive patients admitted to hospital for asthma during an 11-month period were considered for enrollment (n = 497). Those meeting the inclusion criteria were randomized into a control (n = 60) and intervention group (n = 60). Patients in the intervention group received SMS messages according to a structured workflow, while patients in the control group had no SMS support. In the intervention group, the mean response rate to the messages was 82%. There was an improvement in the Asthma Control Test (ACT) scores in 36 subjects in the intervention group compared to 28 subjects in the control group. There were reductions in the number of nebulizations in 54 subjects in the control group compared to 50 subjects in the intervention group, and reductions in emergency department visits in 57 subjects in the control group compared to 51 subjects in the intervention group. However, none of these differences were significant. There was no reduction in admission rates in either group (P = 0.5). The service was accepted by most patients, but its long-term effectiveness on the management of asthma remains to be determined.

Was it Easy to Use an Asthma Control Test (ACT) in Different Clinical Practice Settings in a Tertiary Hospital in Singapore?

Introduction: TheAsthmaControlTestTM (ACT) is a 5-item self-administered tool designed to assess asthma control. It is said to be simple, easy and can be administered quickly by patients in the clinical practice setting. This stated benefit has yet to be demonstrated in our local clinical practice setting. The aim was to identify factors associated with difficulty in the administration of the ACT in different clinical practice settings in a tertiary hospital in Singapore. Materials and Methods: This is a prospective study performed from April to June 2008. All patients diagnosed with asthma and referred to an asthma nurse from the in-patient and out-patient clinical practice setting in Tan Tock Seng Hospital were enrolled. Results: Four hundred and thirty-four patients were asked to complete the ACT tool. In the univariate model, we found that age, clinical setting and medical history to be significantly associated with the completion of the ACT. The odds of completion decreased by a factor of 0.92 (95% CI, 0.89 to 0.94) for every year’s increase in age, and this was statistically significant (P <0.001). Similarly, the odds ratio of completion for those with more than 3 medical conditions by history were 0.59 (95% CI, 0.48 to 0.71) as compared to those with less than 3 medical conditions by history, and this was also significant (P <0.001). In the multivariate model, we only found age to be an independent and significant factor. After adjusting for age, none of the other variables initially significant in the univariate model remained significant. Conclusion: The results show that the ACT was simple and easy to be administered in younger-aged patients. Key words: Administer, Guidelines, Interpretation, Questionnaire

Was it easy to use an Asthma Control Test (ACT) in different clinical practice settings in a tertiary hospital in Singapore?

INTRODUCTION

The Asthma Control Test (ACT) is a 5-item self-administered tool designed to assess asthma control. It is said to be simple, easy and can be administered quickly by patients in the clinical practice setting. This stated benefit has yet to be demonstrated in our local clinical practice setting. The aim was to identify factors associated with difficulty in the administration of the ACT in different clinical practice settings in a tertiary hospital in Singapore.

MATERIALS AND METHODS

This is a prospective study performed from April to June 2008. All patients diagnosed with asthma and referred to an asthma nurse from the in-patient and out-patient clinical practice setting in Tan Tock Seng Hospital were enrolled.

RESULTS

Four hundred and thirty-four patients were asked to complete the ACT tool. In the univariate model, we found that age, clinical setting and medical history to be significantly associated with the completion of the ACT. The odds of completion decreased by a factor of 0.92 (95% CI, 0.89 to 0.94) for every year's increase in age, and this was statistically significant (P <0.001). Similarly, the odds ratio of completion for those with more than 3 medical conditions by history were 0.59 (95% CI, 0.48 to 0.71) as compared to those with less than 3 medical conditions by history, and this was also significant (P <0.001). In the multivariate model, we only found age to be an independent and significant factor. After adjusting for age, none of the other variables initially significant in the univariate model remained significant.

CONCLUSION

The results show that the ACT was simple and easy to be administered in younger-aged patients.