Is predictive coding falsifiable?

Predictive-coding has justifiably become a highly influential theory in Neuroscience. However, the possibility of its unfalsifiability has been raised. We argue that if predictive-coding were unfalsifiable, it would be a problem, but there are patterns of behavioural and neuroimaging data that would stand against predictive-coding. Contra (vanilla) predictive patterns are those in which the more expected stimulus generates the largest evoked-response. However, basic formulations of predictive-coding mandate that an expected stimulus should generate little, if any, prediction error and thus little, if any, evoked-response. It has, though, been argued that contra (vanilla) predictive patterns can be obtained if precision is higher for expected stimuli. Certainly, using precision, one can increase the amplitude of an evoked-response, turning a predictive into a contra (vanilla) predictive pattern. We demonstrate that, while this is true, it does not present an absolute barrier to falsification. This is because increasing precision also reduces latency and increases the frequency of the response. These properties can be used to determine whether precision-weighting in predictive-coding justifiably explains a contra (vanilla) predictive pattern, ensuring that predictive-coding is falsifiable.

Incentive value and spatial certainty combine additively to determine visual priorities.

How does the brain combine information predictive of the value of a visually guided task (incentive value) with information predictive of where task-relevant stimuli may occur (spatial certainty)? Human behavioural evidence indicates that these two predictions may be combined additively to bias visual selection (Additive Hypothesis), whereas neuroeconomic studies posit that they may be multiplicatively combined (Expected Value Hypothesis). We sought to adjudicate between these two alternatives. Participants viewed two coloured placeholders that specified the potential value of correctly identifying an imminent letter target if it appeared in that placeholder. Then, prior to the target's presentation, an endogenous spatial cue was presented indicating the target's more likely location. Spatial cues were parametrically manipulated with regard to the information gained (in bits). Across two experiments, performance was better for targets appearing in high versus low value placeholders and better when targets appeared in validly cued locations. Interestingly, as shown with a Bayesian model selection approach, these effects did not interact, clearly supporting the Additive Hypothesis. Even when conditions were adjusted to increase the optimality of a multiplicative operation, support for it remained. These findings refute recent theories that expected value computations are the singular mechanism driving the deployment of endogenous spatial attention. Instead, incentive value and spatial certainty seem to act independently to influence visual selection.

The experiential blink: Mapping the cost of working memory encoding onto conscious perception in the attentional blink.

The attentional blink (AB) represents a cognitive deficit in reporting the second of two targets (T2), when that second target appears 200-600 msec after the first (T1). However, it is unclear how this paradigm impacts the subjective visibility (that is, the conscious perception) of T2, and whether the temporal profile of T2 report accuracy matches the temporal profile of subjective visibility. In order to compare report accuracy and subjective visibility, we asked participants to identify T1 and T2, and to rate the subjective visibility of T2 across two experiments. Event-related potentials were also measured. The results revealed different profiles for the report of T2 versus the subjective visibility of T2, particularly when T1 and T2 appeared within 200 msec of one another. Specifically, T2 report accuracy was high but T2 visibility was low when the two targets appeared in close temporal succession, suggesting what we call the Experiential Blink is different from the classic AB. Electrophysiologically, at lag-1, the P3 component was modulated more by subjective visibility than by report accuracy. Collectively, the data indicate that the deficit in accurately reporting T2 is not the same as the deficit in subjectively experiencing T2. This suggests that traditional understandings of the AB may require adjustment and that, consistent with other findings, working memory (WM) encoding and conscious perception may not be synonymous.

Latency as a region contrast: Measuring ERP latency differences with Dynamic Time Warping.

Methods for measuring onset latency contrasts are evaluated against a new method utilizing the dynamic time warping (DTW) algorithm. This new method allows latency to be measured across a region instead of single point. We use computer simulations to compare the methods' power and Type I error rates under different scenarios. We perform per-participant analysis for different signal-to-noise ratios and two sizes of window (broad vs. narrow). In addition, the methods are tested in combination with single-participant and jackknife average waveforms for different effect sizes, at the group level. DTW performs better than the other methods, being less sensitive to noise as well as to placement and width of the window selected.

Determination of bromide in canine plasma using ion chromatography.

A new ion chromatographic procedure has been developed and validated for the determination of bromide in canine plasma. Following a simple dilution, samples were separated on a Metrosep A Supp 5 column. The mobile phase was an isocratic mixture of 2.2 mM Na(2)CO(3), 1.0 mM NaHCO(3), and 1% acetonitrile, with a flow-rate of 0.7 ml/min. The procedure produced a linear curve over the concentration range of 50-2500 microg/ml. The development of the assay permitted the determination of therapeutic levels after oral administration of potassium bromide to dogs being treated for epilepsy.

A reciprocal relationship between bottom-up trace strength and the attentional blink bottleneck: relating the LC-NE and ST(2) models.

There is considerable current interest in neural modeling of the attentional blink phenomenon. Two prominent models of this task are the Simultaneous Type Serial Token (ST(2)) model and the Locus Coeruleus-Norepinephrine (LC-NE) model. The former of these generates a broad spectrum of behavioral data, while the latter provides a neurophysiologically detailed account. This paper explores the relationship between these two approaches. Specifically, we consider the spectrum of empirical phenomena that the two models generate, particularly emphasizing the need to generate a reciprocal relationship between bottom-up trace strength and the blink bottleneck. Then we discuss the implications of using ST(2) token mechanisms in the LC-NE setting.

Community elder-care in Tasmania: examining whether caregivers believe they "make-a-difference" in an urban and rural island.

Adult caregivers (n = 184;Mage = 43.9 years old) working at a non-profit, eldercare program at five geographically diverse sites located in the self-contained, island state of Tasmania, Australia, completed a set of self-report measures. Results across the five sites indicated that respondents experienced a relatively strong sense of self-efficacy toward making a difference in their local community. However, there were significant differences (controlling for social desirability) when comparing caregivers from rural northern (n = 45) with urban southern (n = 139) communities, with rural caregivers claiming stronger sense of common mission with others, reciprocal responsibility to help others, and caregiver satisfaction, plus lower disharmony with other members and caregiver stress in helping the elderly than urban caregivers. Implications suggest that community self-efficacy may be high among eldercare staff, but their sense of community and caregiving perceptions may reflect geographic differences, especially in Tasmania.

Thermal diffusion probe analysis of perfusion changes in vascular occlusions of rabbit pedicle flaps.

BACKGROUND: The purpose of this study was to show the efficacy of a novel tissue blood flow measuring device in an animal model. Thermal diffusion technology evaluates changes in perfusion in small volumes of tissue. METHODS: The thermal diffusion probe device is a long, 0.9-mm-diameter flexible catheter with two thermistors, which are placed directly into the tissue; it excites an active thermistor to a constant temperature slightly above the tissue baseline and collects data on the power dissipated in the active thermistor. It also continuously monitors the baseline tissue temperature using an additional passive thermistor placed outside the heated field. In this study, rabbit epigastric pedicle flaps were instrumented with two thermal diffusion probes (peripheral and deep) to continuously monitor flap perfusion. RESULTS: Twenty-five vascular occlusion studies were performed in 16 flaps. Blood vessel occlusions (arterial, venous, and arteriovenous) were easily detectable with this system. Waveforms for arterial and arteriovenous occlusions differed from those for venous occlusions. Probes in both peripheral and deep tissue locations were sensitive to changes in tissue perfusion. CONCLUSION: Thermal diffusion probes may provide a useful clinical method for monitoring flap perfusion.

Continuous and real-time blood perfusion monitoring in prefabricated flaps.

The Thermal Diffusion Probe (TDP) System allows continuous real-time measurement of tissue perfusion in flaps. The authors used a TDP with two thermistors, one active, the other passive, embedded in a 0.9-mm diameter catheter to measure continuous tissue perfusion in rabbit epigastric flaps. The distal thermistor is heated to 2 degrees C above the tissue baseline temperature. The power required to maintain this temperature difference is mathematically related to the tissue perfusion in the volume surrounding the probe tip. Central and peripheral TDPs were placed. The TDP effectiveness in detecting and measuring daily tissue perfusion in buried epigastric flaps was tested. Contralateral epigastric pedicles were transposed into the flaps prior to ligation of the original pedicle. Flaps with transposed pedicles showed a progressive and significant increase in tissue perfusion during the initial 3 weeks of the experiment, compared to flaps without the pedicle transfer. The TDP System is a useful experimental method for the continuous and real-time quantification of flap perfusion and may be helpful in making clinical decisions about prefabricated flap transfer.

Impact of race on the outcome of carotid endarterectomy: a population-based analysis of 9,842 recent elective procedures.

OBJECTIVE: To examine the influence of race and other potentially confounding variables on the outcome of carotid endarterectomy (CEA). SUMMARY: Previous studies have demonstrated that CEA is performed less frequently in black patients, although little attention has been focused on the influence of race on the outcome of surgery. METHODS: The Maryland Health Services Cost Review Commission database was reviewed to identify all elective CEA procedures performed in all nonfederal acute care hospitals in the state from 1990 through 1995 to examine the influence of race and other factors on the rates of in-hospital complications, in-hospital stroke, length of stay, and total hospital charges. RESULTS: Carotid endarterectomy was performed in 9,219 (94%) white and 623 (6%) black patients during this period. The in-hospital stroke rate was 1.7%-3. 1% among black patients and 1.6% among white patients. Black patients had a longer length of stay and higher mean hospital charges than white patients. Multivariate logistic regression analysis identified black race as an independent risk factor for in-hospital stroke. Performance of CEA by a high-volume surgeon was protective for the combined occurrence of in-hospital stroke or death, and whites were more than twice as likely to undergo surgery performed by high-volume surgeons. Conversely, undergoing surgery in a low-volume hospital was associated with in-hospital stroke, and blacks were four times as likely to use low-volume hospitals. CONCLUSIONS: Black patients who underwent elective CEA in Maryland from 1990 to 1995 had an increased incidence of in-hospital stroke, a longer hospital stay, and higher hospital charges than whites. Black race was identified as an independent risk factor for in-hospital stroke, although the reasons for this influence of race on outcome are undefined. The authors' observations also suggest the possibility of limited access to optimal surgical care among blacks, and this issue warrants further study.

Validation of real-time continuous perfusion measurement.

Perfusion, the rate at which blood in tissue is replenished at the capillary level, is a primary factor in the transport of heat, drugs, oxygen and nutrients. While there have been many measurement techniques proposed, most do not lend themselves to routine, continuous and real-time use. A minimally invasive probe, called the thermal diffusion probe (TDP), which uses a self-heated thermistor to measure absolute perfusion continuously and in real time, was validated at low flows with the microsphere technique. In 27 rabbits, simultaneous TDP measurements were made in liver from 0 to 60 ml min-1 100 g-1. The TDP perfusion correlated well with the microspheres (R2 = 0.898) and the agreement between techniques is very good with a slope close to unity (0.921) and an intercept close to zero (0.566 ml min-1 100 g-1). Variability between the two techniques was primarily due to the sampling error from the microsphere 'snap shot' of periodic blood flow when compared with the continuous TDP perfusion measurement. The ability to quantify local perfusion continuously and in real time may have a profound impact on patient management in a number of clinical areas such as organ transplantation, neurosurgery, oncology and others, in which quantitative knowledge of perfusion is of value.

Clinical pathway implementation improves outcomes for complex biliary surgery.

BACKGROUND: Complex biliary surgery is associated with significant morbidity, prolonged hospital stay, and high cost. Clinical pathway implementation has the potential to standardize treatment and improve outcomes. Therefore the aim of this analysis was to determine whether clinical pathway implementation and/or feedback of outcome data would alter hospital stay, charges, and mortality rates for complex biliary surgery at an academic medical center METHODS: Pre- and postoperative length of stay, hospital charges, and mortality rates were monitored for 36 months before (period 1) and for 2 18-month periods (periods 2 and 3) after implementation of a clinical pathway for hepaticojejunostomy. Outcome data were provided to the surgeons 18 months after pathway implementation to determine whether further clinical practice improvement was possible. RESULTS: From 1991 to 1997, 339 patients underwent hepaticojejunostomy at The Johns Hopkins Hospital for malignant and benign biliary obstruction. Total length of stay was 13.3 +/- 0.9 days for period 1 compared with 12.5 +/- 0.8 days for period 2 (not significant) and 10.1 +/- 0.3 days for period 3 (P < .01 vs period 1; P < .03 vs period 2). Hospital charges averaged $24,446 during period 1 compared with $23,338 during period 2 and $20,240 during period 3 (P < .01 vs periods 1 and 2). Hospital mortality rate was 4.5% during period 1 compared with 0.7% during periods 2 and 3 (P < .05). CONCLUSIONS: These data suggest that implementation of a clinical pathway for hepaticojejunostomy reduces hospital mortality rates and that feedback of outcome data to surgeons results in further clinical practice improvement. Thus clinical pathway implementation and feedback are effective methods to control costs at an academic medical center.

Hospital volume can serve as a surrogate for surgeon volume for achieving excellent outcomes in colorectal resection.

OBJECTIVE: To examine the association of surgeon and hospital case volumes with the short-term outcomes of in-hospital death, total hospital charges, and length of stay for resection of colorectal carcinoma. METHODS: The study design was a cross-sectional analysis of all adult patients who underwent resection for colorectal cancer using Maryland state discharge data from 1992 to 1996. Cases were divided into three groups based on annual surgeon case volume--low (< or =5), medium (5 to 10), and high (>10)--and hospital volume--low (<40), medium (40 to 70), and high (> or =70). Poisson and multiple linear regression analyses were used to identify differences in outcomes among volume groups while adjusting for variations in type of resections performed, cancer stage, patient comorbidities, urgency of admission, and patient demographic variables. RESULTS: During the 5-year period, 9739 resections were performed by 812 surgeons at 50 hospitals. The majority of surgeons (81%) and hospitals (58%) were in the low-volume group. The low-volume surgeons operated on 3461 of the 9739 total patients (36%) at an average rate of 1.8 cases per year. Higher surgeon volume was associated with significant improvement in all three outcomes (in-hospital death, length of stay, and cost). Medium-volume surgeons achieved results equivalent to high-volume surgeons when they operated in high- or medium-volume hospitals. CONCLUSIONS: A skewed distribution of case volumes by surgeon was found in this study of patients who underwent resection for large bowel cancer in Maryland. The majority of these surgeons performed very few operations for colorectal cancer per year, whereas a minority performed >10 cases per year. Medium-volume surgeons achieved excellent outcomes similar to high-volume surgeons when operating in medium-volume or high-volume hospitals, but not in low-volume hospitals. The results of low-volume surgeons improved with increasing hospital volume but never equaled those of the high-volume surgeons.

Thermodiffusion for continuous quantification of hepatic microcirculation--validation and potential in liver transplantation.

Hepatic microcirculation is a main determinant of reperfusion injury and graft quality in liver transplantation. Methods available for the quantification of hepatic microcirculation are indirect, are invasive, or preclude postoperative application. The aim of this study was the validation of thermodiffusion in a new modification allowing long-term use in the clinical setting. In six pigs Doppler flowmeters were positioned around the hepatic artery and portal vein for the measurement of total liver blood flow. Liver perfusion was quantified by thermodiffusion and compared to H(2) clearance as an established technique under baseline conditions, during different degrees of portal venous obstruction and during occlusion of the hepatic artery. Thermodiffusion measurements were recorded for five days postoperatively followed by histological evaluation of the hepatic puncture site. Perfusion data obtained by thermodiffusion were significantly correlated to H(2) clearance (r = 0.94, P < 0. 001) and to liver blood flow (r = 0.9, P < 0.05). The agreement between thermodiffusion and H(2) clearance was excellent (mean difference -2.1 ml/100 g/min; limits of agreement -12.5 and 8.3 ml/100 g/min). Occlusion of the portal vein or hepatic artery was immediately detected by thermodiffusion, indicating a decrease of perfusion by 64 +/- 7% or 27 +/- 5% of baseline, respectively. Perfusion values at baseline and during vascular occlusion were reproducible during the entire observation period. Histological changes of the liver tissue adjacent to the thermodiffusion probes were minute and did not influence long-term measurements. In vivo validation proved that enhanced thermodiffusion is a minimally invasive technique for the continuous, real-time quantification of hepatic microcirculation. Changes in liver perfusion can be safely detected over several days postoperatively. The implication for liver transplantation has led to the clinical application of thermodiffusion.

Complex gastrointestinal surgery: impact of provider experience on clinical and economic outcomes.

BACKGROUND: Commonly performed elective gastrointestinal surgical procedures are carried out with low morbidity and mortality in hospitals throughout the United States. Complex operative procedures on the alimentary tract are performed with a relatively low frequency and are associated with higher mortality. Volume and experience of the surgical provider team have been correlated with better clinical and economic outcomes for one complex gastrointestinal surgical procedure, pancreaticoduodenectomy. This study evaluated whether provider volume and experience were important factors influencing clinical and economic outcomes for a variety of complex gastrointestinal surgical procedures in one state. STUDY DESIGN: Complex high-risk gastrointestinal surgical procedures were defined as those with statewide in-hospital mortality of > or = 5%, frequency of greater than 200 per year in the state, and requiring special surgical skill and expertise. Six procedures met these criteria. Using publicly available discharge data, all patients discharged from Maryland hospitals from July 1989 to June 1997 with a primary procedure code for one of the six study procedures were selected. Hospitals were classified into one of six groups based on the average number of study procedures per year: 10 or less; 11 to 20; 21 to 50; 51 to 100; 101 to 200; and 201 or more procedures per year. A hospital was included if at least one procedure was performed there during the study period. No providers fell within the 51 to 100, and 101 to 200 groups, so all analyses were performed for the remaining four volume groups that were classified, respectively, as minimal (10 or fewer procedures), low (11 to 20 procedures), medium (21 to 50 procedures), and high-volume groups (201 or more procedures). Poisson regression was used to assess the relationship between in-hospital mortality and hospital volume after case-mix adjustment. Multiple linear regression models were used to assess differences in average length-of-stay and average total hospital charges among hospital volume groups. We further analyzed mortality, length-of-stay, and charges at the procedural level to understand these subgroups of complex gastrointestinal patients. We also examined the relationship between provider volume and outcomes for malignant versus benign diagnosis groups. RESULTS: Complex gastrointestinal surgical procedures were performed on 4,561 patients in Maryland from July 1989 through June 1997. The study population averaged 61.6 years of age, was 55% male, 71% Caucasian, and had predominantly Medicare as a payment source. After case-mix adjustment, patients who underwent complex gastrointestinal surgical procedures at the medium-, low-, and minimal-volume provider groups had a 2.1, 3.3, and 3.2 times greater risk of in-hospital death, respectively, than patients at the high-volume provider (p < 0.001 for all comparisons); longer lengths-of-stay, 16.1, 15.7, and 15.5 days at the low-, medium-, and minimal-volume groups, respectively, versus 14.0 days for the high-volume provider (p < 0.001 for all comparisons). Similarly, adjusted charges at the high-volume provider were, on average, 14% less than those of the low-volume group, which had the next lowest charges. Although mortality rates differed by procedure type, for each procedure, mortality increased as provider volume decreased, following the pattern found in the aggregate analysis. After case-mix adjustment, the risk of in-hospital death for patients with malignant diagnoses was significantly higher for the medium-, low-, and minimal-volume groups compared with patients at the high-volume provider, relative risk of 3.1, 4.0, and 4.2, respectively, (p < 0.001 for all comparisons). CONCLUSIONS: This study demonstrates that increased hospital experience is associated with a marked decrease in hospital mortality. The decreased mortality at the high-volume provider was also associated with shorter lengths-of-stay and lower hospital char

Cost implications of different surgical management strategies for primary hyperparathyroidism.

BACKGROUND: Controversy exists about optimal management of patients with primary hyperparathyroidism. To date, no studies have explored the cost implications of variation in practice. METHODS: Results from a national survey of endocrine surgeons were combined with results from a survey of endocrinologists and financial data from Medicare. Patterns of use of resources were identified, annual costs for the surgical management of primary hyperparathyroidism in the United States were calculated, and the financial impact of variation in practice was estimated. RESULTS: Survey respondents (n = 109) were experienced endocrine surgeons, performing an average of 33 parathyroidectomies annually. Seventy-five percent of patients undergo localization before initial exploration for primary hyperparathyroidism. In order of preference, these studies were sestamibi (43%), ultrasonography (28%), and sestamibi with single-photon emission computed tomography (26%). Although there is variation in preoperative and postoperative practice, in-hospital costs have the greatest influence on total cost. An estimated $282 million is spent annually in the United States on operations for primary hyperparathyroidism. National health expenditures could range by more than $70 million, depending on whether management strategies involving low or high use of resources are employed. CONCLUSIONS: Substantial variation among endocrine surgeons in the management of primary hyperparathyroidism has important cost implications. Implementation of evidence-based guidelines to optimize clinical and economic performance should be considered.

Should hepatic resections be performed at high-volume referral centers?

Recent studies have demonstrated the relationship between clinical outcomes of complex surgical procedures and provider volume. Hepatic resection is one such high-risk surgical procedure. The aim of this analysis was to determine whether mortality and cost of performing hepatic resection are related to surgical volume while also examining outcomes by extent of resection and diagnosis, variables seen with this procedure. Maryland discharge data were used to study surgical volume, length of stay, charges, and mortality for 606 liver resections performed at all acute-care hospitals between January 1990 and June 1996. One high-volume provider accounted for 43.6% of discharges, averaging 40.6 cases per year. In comparison, the remainder of resections were performed at 35 other hospitals, averaging 1.5 cases per year. Data were stratified into these high- and low-volume groups, and adjusted outcomes were compared. The mortality rate for all procedures in the low-volume group was 7.9% compared to 1.5% for the high-volume provider (P <0.01, relative risk = 5.2). No overall differences were observed between low- and high-volume providers in total hospital charges. When analyzing by procedure type and diagnosis, lower mortality was seen in the high-volume center for both minor and major resections, as well as resections for metastatic disease. It was concluded that hepatic resection can be performed more safely and at comparable cost at high-volume referral centers.

Surgical repair of ruptured abdominal aortic aneurysms in the state of Maryland: factors influencing outcome among 527 recent cases.

PURPOSE: Abdominal aortic aneurysm (AAA) rupture has been historically associated with high operative mortality rates. In this community-based, cross-sectional study, we examined factors influencing outcome after operations performed for ruptured AAA (rAAA). METHODS: An analysis of a state database identified 3820 patients who underwent AAA repair between 1990 and 1995, including 527 (13.8%) who had an operation for an rAAA. Demographic variables examined included patient age, gender, race, associated comorbidity rates, operative surgeon experience with rAAA, and annual hospital rAAA and total AAA operative volumes. Outcomes measured included operative mortality rates, hospital length of stay, and charges. RESULTS: Operative mortality rates increased significantly with advancing age (P < 0.0001) but were not related to gender (P = 0.474) or race (p = 0.598) and were significantly lower among patients with hypertension (P = 0.006) or pulmonary disease (P = 0.045). There was no relationship between hospital rAAA or total AAA volume and rAAA repair mortality rate, although high-volume surgeons (i.e., performing more than 10 rAAA repairs) had decreased mortality rates and hospital charges compared with other surgeons. Hospital lengths of stay and charges increased with age among survivors, but not nonsurvivors, of rAAA repair. Despite a stable incidence of rAAA repairs during the study interval and no significant change in the mean age of patients undergoing operation or the percentage of operations performed by high-volume surgeons, the statewide mortality rate declined from 59.3% to 43.2% (P = 0.039). CONCLUSION: The incidence of rAAA does not appear to be declining. Although operative rAAA repair continues to be associated with substantial risk and remains an especially lethal condition among the elderly, the operative mortality rate has declined in recent years in Maryland. Lower operative mortality rates and hospital charges are associated with operations performed by high-volume surgeons.

The importance of surgeon experience for clinical and economic outcomes from thyroidectomy.

OBJECTIVE: To determine whether individual surgeon experience is associated with improved short-term clinical and economic outcomes for patients with benign and malignant thyroid disease who underwent thyroid procedures in Maryland between 1991 and 1996. SUMMARY BACKGROUND DATA: There is a prevailing belief that surgeon experience affects patient outcomes in endocrine surgery, but there is a paucity of objective evidence outside of clinical series published by experienced surgeons that supports this view. METHODS: A cross-sectional analysis of all patients who underwent thyroidectomy in Maryland between 1991 and 1996 was conducted using a computerized statewide hospital discharge data base. Surgeons were categorized by volume of thyroidectomies over the 6-year study period: A (1 to 9 cases), B (10 to 29 cases), C (30 to 100 cases), and D (>100 cases). Multivariate regression was used to assess the relation between surgeon caseload and in-hospital complications, length of stay, and total hospital charges, adjusting for case mix and hospital volume. RESULTS: The highest-volume surgeons (group D) performed the greatest proportion of total thyroidectomies among the 5860 discharges, and they were more likely to operate on patients with cancer. After adjusting for case mix and hospital volume, highest-volume surgeons had the shortest length of stay (1.4 days vs. 1.7 days for groups B and C and 1.9 days for group A) and the lowest complication rate (5.1 % vs. 6.1% for groups B and C and 8.6% for group A). Length of stay and complications were more determined by surgeon experience than hospital volume, which had no consistent association with outcomes. CONCLUSIONS: Individual surgeon experience is significantly associated with complication rates and length of stay for thyroidectomy.

Importance of hospital volume in the overall management of pancreatic cancer.

OBJECTIVE: To determine whether hospital volume is associated with clinical and economic outcomes for patients with pancreatic cancer who underwent pancreatic resection, palliative bypass, or endoscopic or percutaneous stent procedures in Maryland between 1990 and 1995. SUMMARY BACKGROUND DATA: Previous studies have demonstrated that outcomes for patients undergoing a Whipple procedure improve with higher surgical volume, but only 20% to 35% of patients with pancreatic cancer qualify for curative resection. Most patients undergo palliative procedures instead with a surgical bypass or biliary stent. METHODS: Analysis of hospital discharge data from all nonfederal acute care hospitals in Maryland identified all patients with pancreatic cancer who underwent a pancreatic resection, palliative bypass, or stent procedure between 1990 and 1995. Hospitals (n = 48) were categorized as high-, medium-, and low-volume providers according to their average annual volume of these procedures. Multivariate regression was used to examine the association between hospital volume and in-hospital mortality rate, length of stay, and hospital charges, after adjusting for differences in case mix and surgeon volume. RESULTS: Increased hospital volume is associated with markedly decreased in-hospital mortality rates and a decreased or similar length of stay for all three types of procedures and with decreased or similar hospital charges for resections and stents. After adjustment for case mix differences, the relative risk (RR) of in-hospital death after pancreatic resection was 19.3 and 8 at the low- and medium-volume hospitals, respectively, versus the high-volume hospital; after bypasses, the RR of death was 2.7 and 1.9, respectively; and after stents, the RR was 4.3 and 4.8, respectively. CONCLUSIONS: Patients with pancreatic cancer who are to be treated with curative or palliative procedures appear to benefit from referral to a high-volume provider.