Analyzing the performance of ArcCHECK diode array detector for VMAT plan

Aim

The aim of this study is to evaluate performance of ArcCHECK diode array detector for the volumetric modulated arc therapy (VMAT) patient specific quality assurance (QA). VMAT patient specific QA results were correlated with ion chamber measurement. Dose response of the ArcCHECK detector was studied.

Background

VMAT delivery technique improves the dose distribution. It is complex in nature and requires proper QA before its clinical implementation. ArcCHECK is a novel three dimensional dosimetry system.

Materials and methods

Twelve retrospective VMAT plans were calculated on ArcCHECK phantom. Point dose and dose map were measured simultaneously with ion chamber (IC-15) and ArcCHECK diode array detector, respectively. These measurements were compared with their respective TPS calculated values.

Results

The ion chamber measurements are in good agreement with TPS calculated doses. Mean difference between them is 0.50% with standard deviation of 0.51%. Concordance correlation coefficient (CCC) obtained for ion chamber measurements is 0.9996. These results demonstrate a strong correlation between the absolute dose predicted by our TPS and the measured dose. The CCC between ArcCHECK doses and TPS predictions on the CAX was found to be 0.9978. In gamma analysis of dose map, the mean passing rate was 98.53% for 3% dose difference and 3 mm distance to agreement.

Conclusions

The VMAT patient specific QA with an ion chamber and ArcCHECK phantom are consistent with the TPS calculated dose. Statistically good agreement was observed between ArcCHECK measured and TPS calculated. Hence, it can be used for routine VMAT QA.     

Experimental set up for the irradiation of biological samples and nuclear track detectors with UV C

Aim

In this work we present a methodology to produce an “imprint” of cells cultivated on a polycarbonate detector by exposure of the detector to UV C radiation.

Background

The distribution and concentration of ¹⁰B atoms in tissue samples coming from BNCT (Boron Neutron Capture Therapy) protocols can be determined through the quantification and analysis of the tracks forming its autoradiography image on a nuclear track detector. The location of boron atoms in the cell structure could be known more accurately by the simultaneous observation of the nuclear tracks and the sample image on the detector.

Materials and Methods

A UV C irradiator was constructed. The irradiance was measured along the lamp direction and at different distances. Melanoma cells were cultured on polycarbonate foils, incubated with borophenylalanine, irradiated with thermal neutrons and exposed to UV C radiation. The samples were chemically attacked with a KOH solution.

Results

A uniform irradiation field was established to expose the detector foils to UV C light. Cells could be seeded on the polycarbonate surface. Both imprints from cells and nuclear tracks were obtained after chemical etching.

Conclusions

It is possible to yield cellular imprints in polycarbonate. The nuclear tracks were mostly present inside the cells, indicating a preferential boron uptake.     

Automatic Detection of Whole Night Snoring Events Using Non-Contact Microphone

Objective

Although awareness of sleep disorders is increasing, limited information is available on whole night detection of snoring. Our study aimed to develop and validate a robust, high performance, and sensitive whole-night snore detector based on non-contact technology.

Design

Sounds during polysomnography (PSG) were recorded using a directional condenser microphone placed 1 m above the bed. An AdaBoost classifier was trained and validated on manually labeled snoring and non-snoring acoustic events.

Patients

Sixty-seven subjects (age 52.5±13.5 years, BMI 30.8±4.7 kg/m², m/f 40/27) referred for PSG for obstructive sleep apnea diagnoses were prospectively and consecutively recruited. Twenty-five subjects were used for the design study; the validation study was blindly performed on the remaining forty-two subjects.

Measurements and Results

To train the proposed sound detector, >76,600 acoustic episodes collected in the design study were manually classified by three scorers into snore and non-snore episodes (e.g., bedding noise, coughing, environmental). A feature selection process was applied to select the most discriminative features extracted from time and spectral domains. The average snore/non-snore detection rate (accuracy) for the design group was 98.4% based on a ten-fold cross-validation technique. When tested on the validation group, the average detection rate was 98.2% with sensitivity of 98.0% (snore as a snore) and specificity of 98.3% (noise as noise).

Conclusions

Audio-based features extracted from time and spectral domains can accurately discriminate between snore and non-snore acoustic events. This audio analysis approach enables detection and analysis of snoring sounds from a full night in order to produce quantified measures for objective follow-up of patients.     

MicroPure Imaging for the Evaluation of Microcalcifications in Gouty Arthritis Involving the First Metatarsophalangeal Joint: A Preliminary Study

Objective

To assess the value of MicroPure, a new ultrasound image processing technique, in identifying microcalcifications (formed by monosodium urate crystals) in the first metatarsophalangeal joints attacked by gout compared to gray-scale ultrasound images.

Materials and Methods

Thirty-six patients who fulfilled the study inclusion criteria underwent gray-scale ultrasound and MicroPure examinations of the first metatarsophalangeal joints attacked by gout. Static images of the target areas were acquired using gray-scale ultrasound and MicroPure. Two independent and blinded investigators analyzed the images to determine the number of microcalcifications and to score for image quality and artifacts.

Results

The two investigators observed significantly more microcalcifications with MicroPure compared to gray-scale ultrasound (ρ<0.001). The level of agreement between the investigators consistently increased from gray-scale ultrasound to MicroPure imaging (gray-scale interclass correlation coefficient of 0.69 vs. MicroPure interclass correlation coefficient of 0.81). One investigator preferred the MicroPure image quality over gray-scale ultrasound (ρ<0.001), but the other investigator disagreed (ρ<0.001). Both investigators observed fewer artifacts with MicroPure than with gray-scale ultrasound (ρ<0.009).

Conclusion

MicroPure imaging identified significantly more microcalcifications than gray-scale ultrasound.     

A Robust Post-Processing Workflow for Datasets with Motion Artifacts in Diffusion Kurtosis Imaging

Purpose

The aim of this study was to develop a robust post-processing workflow for motion-corrupted datasets in diffusion kurtosis imaging (DKI).

Materials and methods

The proposed workflow consisted of brain extraction, rigid registration, distortion correction, artifacts rejection, spatial smoothing and tensor estimation. Rigid registration was utilized to correct misalignments. Motion artifacts were rejected by using local Pearson correlation coefficient (LPCC). The performance of LPCC in characterizing relative differences between artifacts and artifact-free images was compared with that of the conventional correlation coefficient in 10 randomly selected DKI datasets. The influence of rejected artifacts with information of gradient directions and b values for the parameter estimation was investigated by using mean square error (MSE). The variance of noise was used as the criterion for MSEs. The clinical practicality of the proposed workflow was evaluated by the image quality and measurements in regions of interest on 36 DKI datasets, including 18 artifact-free (18 pediatric subjects) and 18 motion-corrupted datasets (15 pediatric subjects and 3 essential tremor patients).

Results

The relative difference between artifacts and artifact-free images calculated by LPCC was larger than that of the conventional correlation coefficient (p<0.05). It indicated that LPCC was more sensitive in detecting motion artifacts. MSEs of all derived parameters from the reserved data after the artifacts rejection were smaller than the variance of the noise. It suggested that influence of rejected artifacts was less than influence of noise on the precision of derived parameters. The proposed workflow improved the image quality and reduced the measurement biases significantly on motion-corrupted datasets (p<0.05).

Conclusion

The proposed post-processing workflow was reliable to improve the image quality and the measurement precision of the derived parameters on motion-corrupted DKI datasets. The workflow provided an effective post-processing method for clinical applications of DKI in subjects with involuntary movements.     

Differences in Survival between Colon and Rectal Cancer from SEER Data

Background

Little is known about colorectal cancer or colon and rectal cancer. Are they the same disease or different diseases?

Objectives

The aim of this epidemiology study was to compare the features of colon and rectal cancer by using recent national cancer surveillance data.

Design and setting

Data included colorectal cancer (1995–2008) from the Surveillance, Epidemiology, and End Results Program (SEER) database. Only adenocarcinoma was included for analysis.

Patients

A total of 372,130 patients with a median follow-up of 32 months were analyzed.

Main outcome measures

Mean survival of patients with the same stage of colon and rectal cancer was evaluated.

Results

Around 35% of patients had stage information. Among them, colon cancer patients had better survival than those with rectal cancer, by a margin of 4 months in stage IIB. In stage IIIC and stage IV, rectal cancer patients had better survival than colon cancer patients, by about 3 months. Stage IIB colorectal cancer patients had a poorer prognosis than those with stage IIIA and IIIB colorectal cancer. After adjustment of age, sex and race, colon cancer patients had better survival than rectal cancer of stage IIB, but in stage IIIC and IV, rectal cancer patients had better survival than colon cancer.

Limitations

The study is limited by its retrospective nature.

Conclusion

This was a population-based study. The prognosis of rectal cancer was not worse than that of colon cancer. Local advanced colorectal cancer had a poorer prognosis than local regional lymph node metastasis. Stage IIB might require more aggressive chemotherapy, and no less than that for stage III.     

Study on measuring device arrangement of array-type CdTe detector for BNCT-SPECT

Aim

To design the measuring device arrangement of array-type CdTe detector for BNCT-SPECT.

Background

In a boron neutron capture therapy, a very serious unsolved problem exists, namely that the treatment effect for BNCT cannot be known during irradiation in real time. Therefore, we have been developing a so-called BNCT-SPECT with a CdTe detector, which can obtain a three-dimensional image for the BNCT treatment effect by measuring 478 keV gamma-rays emitted from the excited state of ⁷Li nucleus created by the ¹⁰B(n,α) reaction. However, no practical uses were realized at present, because BNCT-SPECT requires very severe conditions for spatial resolution, measuring time, statistical accuracy and energy resolution.

Materials and methods

The design study was performed with numerical simulations carried out by a 3-dimenaional transport code, MCNP5 considering the detector assembly, irradiation room and even arrangement of arrayed CdTe crystals.

Results

The estimated count rate of 478 keV gamma-rays was sufficiently large being more than the target value of over 1000 counts/h. However, the S/N ratio did not meet the target of S/N > 1. We confirmed that deterioration of the S/N ratio was caused by the influence of Compton scattering especially due to capture gamma-rays of hydrogen. Theoretical calculations were thereafter carried out to find out whether anti-Compton measurement in an array-type CdTe detector could decrease the noise due to Compton scatterings.

Conclusions

The calculation result showed that the anti-coincidence would possibly increase the S/N ratio. In the next phase, an arrayed detector with two CdTe crystals will be produced to test removal possibility of the anti-coincident event.     

An Augmented Lagrangian Based Compressed Sensing Reconstruction for Non-Cartesian Magnetic Resonance Imaging without Gridding and Regridding at Every Iteration

Background

Non-Cartesian trajectories are used in a variety of fast imaging applications, due to the incoherent image domain artifacts they create when undersampled. While the gridding technique is commonly utilized for reconstruction, the incoherent artifacts may be further removed using compressed sensing (CS). CS reconstruction is typically done using conjugate-gradient (CG) type algorithms, which require gridding and regridding to be performed at every iteration. This leads to a large computational overhead that hinders its applicability.

Methods

We sought to develop an alternative method for CS reconstruction that only requires two gridding and one regridding operation in total, irrespective of the number of iterations. This proposed technique is evaluated on phantom images and whole-heart coronary MRI acquired using 3D radial trajectories, and compared to conventional CS reconstruction using CG algorithms in terms of quantitative vessel sharpness, vessel length, computation time, and convergence rate.

Results

Both CS reconstructions result in similar vessel length (P = 0.30) and vessel sharpness (P = 0.62). The per-iteration complexity of the proposed technique is approximately 3-fold lower than the conventional CS reconstruction (17.55 vs. 52.48 seconds in C++). Furthermore, for in-vivo datasets, the convergence rate of the proposed technique is faster (60±13 vs. 455±320 iterations) leading to a ∼23-fold reduction in reconstruction time.

Conclusions

The proposed reconstruction provides images of similar quality to the conventional CS technique in terms of removing artifacts, but at a much lower computational complexity.     

All That Glitters Isn't Gold: A Survey on Acknowledgment of Limitations in Biomedical Studies

Background

Acknowledgment of all serious limitations to research evidence is important for patient care and scientific progress. Formal research on how biomedical authors acknowledge limitations is scarce.

Objectives

To assess the extent to which limitations are acknowledged in biomedical publications explicitly, and implicitly by investigating the use of phrases that express uncertainty, so-called hedges; to assess the association between industry support and the extent of hedging.

Design

We analyzed reporting of limitations and use of hedges in 300 biomedical publications published in 30 high and medium -ranked journals in 2007. Hedges were assessed using linguistic software that assigned weights between 1 and 5 to each expression of uncertainty.

Results

Twenty-seven percent of publications (81/300) did not mention any limitations, while 73% acknowledged a median of 3 (range 1–8) limitations. Five percent mentioned a limitation in the abstract. After controlling for confounders, publications on industry-supported studies used significantly fewer hedges than publications not so supported (p = 0.028).

Limitations

Detection and classification of limitations was – to some extent – subjective. The weighting scheme used by the hedging detection software has subjective elements.

Conclusions

Reporting of limitations in biomedical publications is probably very incomplete. Transparent reporting of limitations may protect clinicians and guideline committees against overly confident beliefs and decisions and support scientific progress through better design, conduct or analysis of new studies.     

Prospective Study of the Quality of Colonoscopies Performed by Primary Care Physicians: The Alberta Primary Care Endoscopy (APC-Endo) Study

Background

The quality of colonoscopies performed by primary care physicians (PCPs) is unknown.

Objective

To determine whether PCP colonoscopists achieve colonoscopy quality benchmarks, and patient satisfaction with having their colonoscopy performed by a primary care physician.

Design

Prospective multi-center, multi-physician observational study. Colonoscopic quality data collection occurred via completion of case report forms and pathological confirmation of lesions. Patient satisfaction was captured by a telephone survey.

Setting

Thirteen rural and suburban hospitals in Alberta, Canada.

Measurements

Proportion of successful cecal intubations, average number of adenomas detected per colonoscopy, proportion of patients with at least one adenoma, and serious adverse event rates; patient satisfaction with their wait time and procedure, as well as willingness to have a repeat colonoscopy performed by their primary care endoscopist.

Results

In the two-month study period, 10 study physicians performed 577 colonoscopies. The overall adjusted proportion of successful cecal intubations was 96.5% (95% CI 94.6–97.8), and all physicians achieved the adjusted cecal intubation target of ≥90%. The average number of ademonas detected per colonoscopy was 0.62 (95% CI 0.5–0.74). 46.4% (95% CI 38.5–54.3) of males and 30.2% (95% CI 22.3–38.2) of females ≥50 years of age having their first colonoscopy, had at least one adenoma. Four serious adverse events occurred (three post polypectomy bleeds and one perforation) and 99.3% of patients were willing to have a repeat colonoscopy performed by their primary care colonoscopist.

Limitations

Two-month study length and non-universal participation by Alberta primary care endoscopists.

Conclusions

Primary care physician colonoscopists can achieve quality benchmarks in colonoscopy. Training additional primary care physicians in endoscopy may improve patient access and decrease endoscopic wait times, especially in rural settings.     

Separating Drought Effects from Roof Artifacts on Ecosystem Processes in a Grassland Drought Experiment

1

Given the predictions of increased drought probabilities under various climate change scenarios, there have been numerous experimental field studies simulating drought using transparent roofs in different ecosystems and regions. Such roofs may, however, have unknown side effects, called artifacts, on the measured variables potentially confounding the experimental results. A roofed control allows the quantification of potential artifacts, which is lacking in most experiments.

2

We conducted a drought experiment in experimental grasslands to study artifacts of transparent roofs and the resulting effects of artifacts on ecosystems relative to drought on three response variables (aboveground biomass, litter decomposition and plant metabolite profiles). We established three drought treatments, using (1) transparent roofs to exclude rainfall, (2) an unroofed control treatment receiving natural rainfall and (3) a roofed control, nested in the drought treatment but with rain water reapplied according to ambient conditions.

3

Roofs had a slight impact on air (+0.14°C during night) and soil temperatures (−0.45°C on warm days, +0.25°C on cold nights), while photosynthetically active radiation was decreased significantly (−16%). Aboveground plant community biomass was reduced in the drought treatment (−41%), but there was no significant difference between the roofed and unroofed control, i.e., there were no measurable roof artifact effects.

4

Compared to the unroofed control, litter decomposition was decreased significantly both in the drought treatment (−26%) and in the roofed control treatment (−18%), suggesting artifact effects of the transparent roofs. Moreover, aboveground metabolite profiles in the model plant species Medicago x varia were different from the unroofed control in both the drought and roofed control treatments, and roof artifact effects were of comparable magnitude as drought effects.

5

Our results stress the need for roofed control treatments when using transparent roofs for studying drought effects, because roofs can cause significant side effects.     

Whole abdominal radiotherapy in ovarian cancer

Objectives

The aim of the study was to evaluate the clinical outcome and toxicity after adjuvant whole abdominal radiotherapy (WART) in patients with ovarian cancer.

Material and methods

Ten patients with optimal cytoreduced ovarian cancer, with a mean age of 58 years (40–70) and stage Ic: 4, stage II: 2, stage III: 4, were treated with WART and adjuvant chemotherapy (9/10). The total radiation dose was 22.5 Gy in the whole abdomen and 42–45 Gy in the pelvis.

Results

The mean follow-up was 8 years. The 5-year actuarial disease-free survival (DFS) was 60%, and the overall survival (OS) was 70%. Four patients had disease recurrence. The sites of recurrence were the abdomen in 2 patients and distant metastases in the other 2 patients (liver and brain metastasis). Gastrointestinal toxicity was as follows: acute 3/10 grades I and II, and late toxicity: 2/10 grades I and II, and only 1 patient developed small bowel obstruction (SBO) that required surgery.

Conclusions

Whole abdominal radiotherapy after surgery and platinum-based chemotherapy achieves high locoregional disease control with an acceptable risk of acute toxicity.     

Treatment of pure uterine sarcoma at the Institut Català D’Oncologia

Aim

The aim of this retrospective study was to investigate the clinical and histopathological characteristics of the disease and treatment outcome of patients with pure uterine sarcomas.

Background

Uterine sarcomas are especially rare tumours, comprising only 3–5% of uterine cancers. They are characterized by histopathological diversity, rapid clinical progression, and poor prognosis. Optimal management consists of complete surgical removal and adjuvant radiotherapy may improve the prognosis.

Materials and methods

All patients with pure uterine sarcoma histology treated at our centre, the Institut Català D’Oncologia in Barcelona Spain, between 2002 and 2010 were reviewed.

Results

Records of 17 patients treated at our hospital over an 8-year period were obtained. Nine patients (53%) had leiomyosarcoma, 7 (41%) had endometrial stromal sarcoma, and 1 patient had unclassified sarcoma. All patients were treated with external beam radiation after surgical excision. Mean age was 62 years (range, 51–69 years). Of the 17 patients, 13 (76%) presented with stage I disease, 2 (12%) were stage II, and 2 (12%) stage III. The overall actuarial 2-year survival estimate was 82.5%. Two patients experienced local relapse. The 2-year local control rate was 90%. A total of 5 patients experienced either local or metastatic relapse. The 2-year progression free survival rate was 58%.

Conclusion

In our experience, combined treatment (surgery and adjuvant radiation therapy) is effective with acceptable side effects. Larger and multicenter studies are needed to assess treatment outcome for pure uterine sarcoma histology.     

Initial medical attention on patients with early-stage non-small cell lung cancer

Background

Detection of early stage non-small cell lung cancer (NSCLC) is commonly believed to be incidental. Understanding the reasons that caused initial detection of these patients is important for early diagnosis. However, these reasons are not well studied.

Methods

We retrospectively reviewed medical records of patients diagnosed with stage I or II NSCLC between 2000 and 2009 at UT MD Anderson Cancer Center. Information on suggestive LC-symptoms or other reasons that caused detection were extracted from patients'' medical records. We applied univariate and multivariate analyses to evaluate the association of suggestive LC-symptoms with tumor size and patient survival.

Results

Of the 1396 early stage LC patients, 733 (52.5%) presented with suggestive LC-symptoms as chief complaint. 347 (24.9%) and 287 (20.6%) were diagnosed because of regular check-ups and evaluations for other diseases, respectively. The proportion of suggestive LC-symptom-caused detection had a linear relationship with the tumor size (correlation 0.96; with p<.0001). After age, gender, race, smoking status, therapy, and stage adjustment, the symptom-caused detection showed no significant difference in overall and LC-specific survival when compared with the other (non-symptom-caused) detection.

Conclusion

Symptoms suggestive of LC are the number one reason that led to detection in early NSCLC. They were also associated with tumor size at diagnosis, suggesting early stage LC patients are developing symptoms. Presence of symptoms in early stages did not compromise survival. A symptom-based alerting system or guidelines may be worth of further study to benefit NSCLC high risk individuals.     

Artifacts in magnetic resonance imaging and computed tomography caused by dental materials

Background

Artifacts caused by dental restorations, such as dental crowns, dental fillings and orthodontic appliances, are a common problem in MRI and CT scans of the head and neck. The aim of this in-vitro study was to identify and evaluate the artifacts produced by different dental restoration materials in CT and MRI images.

Methods

Test samples of 44 materials (Metal and Non-Metal) commonly used in dental restorations were fabricated and embedded with reference specimens in gelatin moulds. MRI imaging of 1.5T and CT scan were performed on the samples and evaluated in two dimensions. Artifact size and distortions were measured using a digital image analysis software.

Results

In MRI, 13 out of 44 materials produced artifacts, while in CT 41 out of 44 materials showed artifacts. Artifacts produced in both MRI and CT images were categorized according to the size of the artifact.

Significance

Metal based restoration materials had strong influence on CT and less artifacts in MRI images. Rare earth elements such as Ytterbium trifluoride found in composites caused artifacts in both MRI and CT. Recognizing these findings would help dental materials manufacturers and developers to produce materials which can cause less artifacts in MRI and CT images.     

Likelihood and Predictors of ST-Elevation in Patients Hospitalized for Myocardial Infarction

Importance

Emergency treatment options in myocardial infarction are guided by presence or absence of ST-elevations in electrocardiography. Occurrence and factors associated with ST-presentation in different population groups are however inadequately known.

Objective

To determine likelihood and patient features associated with ST-elevations in myocardial infarction.

Design

Nationwide registry study including 22 hospitals with angiolaboratory during an eight year period in Finland.

Setting

Hospitalized care.

Participants

68,162 consecutive patients aged ≥30 with myocardial infarction.

Measures

Likelihood and patient features associated with presence of ST-elevations.

Results

Myocardial infarction presented with ST-elevation in 37.5% (CI 37.0–37.9%) and without in 62.5% (CI 61.9–63.1%) of patients, p<0.0001. Majority of patients aged 30–59 years with myocardial infarction had ST-elevation, but among octogenarians ST-elevations were present in only 24.7%. Presence of ST-elevations decreased with age by estimated 15.6% (CI 15.0–16.2%) per 10 year increase (p<0.0001). Men aged 40–79 years had significantly higher rate for ST-elevation myocardial infarction compared to women. Sex-based difference in presentation of myocardial infarction declined with increasing age. Overall, men had a 13% (CI 11–15%, p<0.0001) higher relative risk for ST-elevations compared to women when adjusted for age and co-morbidities. Diabetes, atrial fibrillation, peripheral or cerebral artery disease, chronic pulmonary disease, malignancy, and renal insufficiency were associated with absence of ST-elevations in myocardial infarction in multivariate analysis.

Conclusions and Relevance

Myocardial infarction presents with ST-elevations more commonly in men. Presence of ST-elevations decreases with increasing age. Diabetes, atrial fibrillation, peripheral or cerebral artery disease, chronic pulmonary disease, malignancy, and renal insufficiency are associated with absence of ST-elevations in myocardial infarction. These findings may help to predict likelihood of ST-elevations in a patient with myocardial infarction.     

The Effect of Hospital Volume on Mortality in Patients Admitted with Severe Sepsis

Importance

The association between hospital volume and inpatient mortality for severe sepsis is unclear.

Objective

To assess the effect of severe sepsis case volume and inpatient mortality.

Design Setting and Participants

Retrospective cohort study from 646,988 patient discharges with severe sepsis from 3,487 hospitals in the Nationwide Inpatient Sample from 2002 to 2011.

Exposures

The exposure of interest was the mean yearly sepsis case volume per hospital divided into tertiles.

Main Outcomes and Measures

Inpatient mortality.

Results

Compared with the highest tertile of severe sepsis volume (>60 cases per year), the odds ratio for inpatient mortality among persons admitted to hospitals in the lowest tertile (≤10 severe sepsis cases per year) was 1.188 (95% CI: 1.074–1.315), while the odds ratio was 1.090 (95% CI: 1.031–1.152) for patients admitted to hospitals in the middle tertile. Similarly, improved survival was seen across the tertiles with an adjusted inpatient mortality incidence of 35.81 (95% CI: 33.64–38.03) for hospitals with the lowest volume of severe sepsis cases and a drop to 32.07 (95% CI: 31.51–32.64) for hospitals with the highest volume.

Conclusions and Relevance

We demonstrate an association between a higher severe sepsis case volume and decreased mortality. The need for a systems-based approach for improved outcomes may require a high volume of severely septic patients.