Automatic kernel regression modelling using combined leave-one-out test score and regularised orthogonal least squares.

This paper introduces an automatic robust nonlinear identification algorithm using the leave-one-out test score also known as the PRESS (Predicted REsidual Sums of Squares) statistic and regularised orthogonal least squares. The proposed algorithm aims to achieve maximised model robustness via two effective and complementary approaches, parameter regularisation via ridge regression and model optimal generalisation structure selection. The major contributions are to derive the PRESS error in a regularised orthogonal weight model, develop an efficient recursive computation formula for PRESS errors in the regularised orthogonal least squares forward regression framework and hence construct a model with a good generalisation property. Based on the properties of the PRESS statistic the proposed algorithm can achieve a fully automated model construction procedure without resort to any other validation data set for model evaluation.

A robust nonlinear identification algorithm using PRESS statistic and forward regression.

This paper introduces a new robust nonlinear identification algorithm using the predicted residual sums of squares (PRESS) statistic and forward regression. The major contribution is to compute the PRESS statistic within a framework of a forward orthogonalization process and hence construct a model with a good generalization property. Based on the properties of the PRESS statistic the proposed algorithm can achieve a fully automated procedure without resort to any other validation data set for iterative model evaluation.

Improving outcomes for patients hospitalized with CHF.

Congestive Heart Failure is a costly debilitating medical condition that affects millions of elderly people. This Clinical Practice Improvement study was under taken to determine how variation in patient characteristics and clinical processes impact patient outcomes including length of stay, total charges, and increased severity of illness during hospitalization. The AHCPR Guideline for Heart Failure was used to evaluate clinical care. Data on patient characteristics including severity of illness and process of care were analyzed to determine which variables had the greatest impact on outcomes of care. Results showed that patient characteristics including comorbidities, severity of illness, and noncompliance with diet were significantly related to longer lengths of stay, higher charges, and/or increased severity (all covariates, p < .05). In addition, process of care variables including medication use and patient education significantly (p < .05) impacted outcomes of care. Findings have implications for the development of practice guidelines designed to improve patient outcomes associated with CHF hospitalizations.

Soft tissue injuries with the use of safe corridors for transfixion wire placement during external fixation of distal tibia fractures: an anatomic study.

OBJECTIVES: To determine which soft tissue structures are at risk and when joint violation can occur during small wire placement for hybrid external fixation of distal tibial fractures while adhering to published guidelines. DESIGN: Cadaver anatomic experiment. SETTING: University orthopaedic program. SUJBECTS: Five embalmed cadavers. INTERVENTION: Placement of small wire transfixion pins in the distal tibia. MAIN OUTCOME MEASUREMENTS: Dissection and measurements. METHODS: Four orthopaedic surgeons were shown diagrams that have been widely accepted as allowing for placement of transfixion pins in the distal tibia through safe corridors. Each of the orthopaedic surgeons was then asked to place two transfixion pins into each of five cadaver legs in a position that would provide stable external fixation of the metaphysis to the diaphysis with a circular fixator (forty pins total) for a distal tibial fracture within five centimeters of the plafond. The specimens were dissected, and pins impaling neurovascular structures, tendons, or the ankle capsule were recorded. The superior capsular synovial reflections were measured from the anterior joint line and the tip of the medial malleolus. These measurements were also performed on arthrograms of two extremities before their dissection. RESULTS: Fifty-five percent of the pins placed impaled at least one tendon that crosses the ankle joint. Neurovascular structures that were impaled included the saphenous vein (+/-10.5 percent) and the superficial peroneal nerve (+/-7.5 percent). One pin violated the superior capsular synovial reflection, which was an average of thirty-two millimeters (+/-1.58 millimeters) from the tip of the medial malleolus and twenty-one millimeters (+/-1.63 millimeters) from the anteromedial joint line. CONCLUSIONS: This study shows that tendons and neurovascular structures above the ankle are at risk during small transfixion pin placement, even when using safe corridors. Pins placed within two centimeters of the anterior joint line or three centimeters from the medial malleolus may be intracapsular.

Fixed minidose versus-adjusted low-dose warfarin after total joint arthroplasty: a randomized prospective study.

Widespread use of adjusted low-dose warfarin has been limited by the inconvenience of outpatient laboratory monitoring and the perceived risk of bleeding complications. We sought to determine if the dose of warfarin could be lowered safely even further, eliminating the need for laboratory monitoring and lowering the complication rate. Two hundred forty-five Patients undergoing primary total joint arthroplasty (n = 245) were randomized prospectively to adjusted low-dose warfarin (international normalized ratio [INR], 1.4-1.8) or fixed minidose warfarin (2 mg daily, regardless of INR) before hospital discharge. Prophylaxis continued for 6 weeks, with twice-weekly laboratory monitoring. Patients were followed for bleeding, thromboembolic events, and minor reported complications of warfarin therapy. With the numbers available, the rates of thromboembolic and bleeding events were not significantly different using equivalence analysis. Of patients in the fixed group, 8% had INRs >3.1, necessitating a decrease in dosage to 1 mg. Although such a fixed-dose protocol may simplify outpatient prophylaxis, intermittent monitoring still would be required because a subset of patients achieve a moderate level of anticoagulation and would be at risk for bleeding complications.

Newness of drugs and use of HMO services by asthma patients.

OBJECTIVE: To define newness of drug technology and show associations between two measures of newness and health service utilization. METHODS: Healthcare use and changes in severity at each office visit were assessed for 1309 asthma patients from six health maintenance organizations (HMOs) during 1992. The age of each drug product, derived by subtracting its Food and Drug Administration (FDA) approval date from January 1, 1992, was used to construct two newness measures: the average age of all asthma drugs and, separately, all non-asthma drugs a patient used during the year and the percentages of a patient's asthma drugs from each of four time intervals of asthma drug breakthroughs. Service utilization variables included all primary care provider (PCP) visits, total prescription costs, emergency department (ED) visits, and hospitalizations. RESULTS: Using either measure of drug newness, multivariate analyses showed an association between greater use of newer asthma drugs and lower overall drug costs and fewer PCP visits. A trend was found between greater use of newer asthma drugs and fewer hospitalizations and ED visits. Newer non-asthma medications were associated with fewer ED visits. CONCLUSIONS: After controlling for patient and site variables, greater use of newer asthma drugs was associated with significantly lower drug costs and fewer PCP visits; associations with hospitalization rates and ED visits, although lower, were not significant.

Effect of femoral component offset on polyethylene wear in total hip arthroplasty.

Seventeen staged, bilateral total hip arthroplasties performed in 17 patients were reviewed to compare side-to-side polyethylene wear. Implants used on both sides were similar except for implant offset: one hip in each patient was replaced using a femoral component having a standard implant offset, whereas the other side had a lateral offset implant. The mean followup was 5.70 years (range, 2-10.2 years) on the side with a standard femoral implant and 5.67 years (range, 2-9.7 years) on the side with a lateralized femoral component. The only statistically different parameter between the sides was the femoral component offset. All other parameters affecting polyethylene wear, such as period of followup, head size, head type, cup size, cup inclination, medialization of cup, and patient-related factors were similar on both sides. On the side with a standard femoral component, the mean actual prosthetic offset (determined by manufacturer's specifications) was 35.2 mm and the radiologic offset was 31.5 mm. On the side with a lateralized femoral component, the actual prosthetic offset was 42.5 mm and the radiologic offset was 40.1 mm. The difference in offsets between the sides was statistically significant. The mean preoperative offset of the femur was 38.8 mm. Regression analysis revealed that only femoral component offset and cup size correlated significantly with linear wear rate. On the side with a standard femoral component, the linear wear rate was 0.21 mm per year, whereas on the side with a lateralized femoral component, the linear wear rate was 0.10 mm per year. The differences in the linear wear rates were significant. Lateralization of the femoral component in this series more closely restored preoperative hip biomechanics and significantly decreased polyethylene wear.

Cementless femoral fixation in the rheumatoid patient undergoing total hip arthroplasty: minimum 5-year results.

Sixty-two total hip arthroplasties in 49 patients with a diagnosis of rheumatoid arthritis were performed between November 1986 and December 1992. All components were titanium alloy with a circumferential plasma-spray porous coating. Four patients (4 hips) died before 5-year follow-up, and 6 patients (8 hips) were lost to follow-up, leaving 39 patients (50 hips) for review at a minimum 5-year follow-up after surgery (mean, 8 years; range, 5-12 years). There were 12 men and 27 women, with a mean age at time of surgery of 55 years (range, 25-77 years) and a mean weight of 69 kg (range, 42-109 kg). Compared with the preoperative Charnley scores, there was significant improvement in the postoperative scores: pain, from 2.7 to 5.7, and function, from 3.2 to 5.3. Thigh pain was present in 1 patient (1 hip) (2.0%). No femoral fractures occurred intraoperatively with the insertion of the prosthesis. Spot welds consistent with bone ingrowth were identified in all of the femoral components. No femoral components showed evidence of radiographic loosening or required revision for aseptic loosening or incapacitating thigh pain, but 7 acetabular revisions were performed. Uncemented femoral fixation with this component design in rheumatoid patients appears to be a promising treatment.

Major trauma in elderly adults receiving lipid-lowering medications.

BACKGROUND: Some clinical trials, laboratory experiments, and in vitro studies suggest that lipid-lowering medications predispose a person to traumatic injury. METHODS: We used population-based administrative database analysis to study adults age 65 years or more over a 5-year interval (n = 1,348,259). RESULTS: About 12% of the cohort received a prescription for a lipid-lowering medication and about 88% did not. The two groups had similar distributions of age, gender, and income. Overall, 2,557 (0.2%) were hospitalized for major trauma. Those who received a lipid-lowering medication were 39% less likely to sustain a major trauma than those who did not receive such medication (95% confidence interval, 29 to 47). Similar results were observed after adjustment for age, gender, and income; cardiac and neurologic medications; and lethality. No other cardiac or neurologic medication was associated with an apparent safety advantage. CONCLUSION: Lipid-lowering medications do not lead to a clinically important increase in the absolute risk of major trauma for elderly patients in the community.

Primary cementless total hip arthroplasty in octogenarians. Two to eleven-year follow-up.

BACKGROUND: Cementless total hip arthroplasty is an accepted alternative to total hip arthroplasty with cement in younger patients, but it remains controversial for elderly patients. The purpose of this study was to evaluate the clinical and radiographic outcomes of cementless total hip arthroplasty with use of a proximally coated stem in patients who were at least eighty years of age at the time of the operation. METHODS: One hundred and twenty-three cementless total hip replacements were performed for the treatment of osteoarthritis in 114 patients between the ages of eighty and eighty-nine years. Seven patients (eight hips) died within two years after the surgery, seventeen patients (eighteen hips) died more than two years postoperatively but were not followed for at least two years, and five hips were lost to follow-up; this left ninety-two hips in eighty-six patients for review. The mean duration of follow-up was five years (range, two to eleven years). For the clinical evaluation, the Charnley modification of the Merle d'Aubigné and Postel scale was used. In addition, preoperative and postoperative Harris hip scores were available for sixty-nine hips. Seventy-eight hips were followed radiographically for two years or more. The focus of the radiographic evaluation was the status of the fixation of the femoral and acetabular components as well as cup wear. RESULTS: Perioperative medical complications occurred in association with 24% (thirty) of the 123 operations, but there were no deaths. The mean Charnley scores for pain and function for the ninety-two hips that were followed clinically for at least two years improved by 3.0 and 1.4 points, respectively. The sixty-nine hips for which preoperative and postoperative Harris hip scores were available had a mean improvement of 42 points, with a mean score of 82 points at the last follow-up evaluation. Mild thigh pain was present in four patients, but it did not limit their activity. There were no femoral component revisions. All of the femoral components were radiographically stable and had bone ingrowth. No acetabular component failed by loosening, but 41% (thirty) of the seventy-three hips with radiographs available for measurement of wear showed polyethylene wear. Of the seventy-eight cups that were followed radiographically for two years or more, 4% (three) were associated with lysis, but none had been revised. CONCLUSIONS: Cementless fixation in the elderly is safe, effective, and durable at the time of two to eleven-year follow-up.

Total hip arthroplasty using two different cementless tapered stems.

The authors report their 15-year experience with primary total hip arthroplasty using collarless, tapered, porous-coated femoral stems (Trilock and Taperloc) in patients with osteoarthritis, rheumatoid arthritis, and in octogenarians. Excellent clinical results were achieved in all groups at latest followup. For the patients with Trilock stems, Taperloc stems, and patients who were octogenarians and patients with rheumatoid arthritis, Charnley pain scores were 5.6, 5.5, 5.7, and 5.7; Charnley function scores were 5.2, 5.1, 4.2, and 5.3; Harris hip scores were 92, 92, 82, and 93 points, respectively. There was a 2% rate of thigh pain with the Trilock, 4% with Taperloc, 4% in octogenarians, and 2% in patients with rheumatoid arthritis. In 96% of the patients in the Trilock group, in 100% of the patients in the Taperloc group, in 100% of the patients who were octogenarians, and in 100% of the patients with rheumatoid arthritis, femoral components showed radiographic evidence of bone ingrowth. There were six (12%) femoral component revisions in the Trilock group (all secondary to nonmodularity of the component at the time of acetabular revision), one femoral component revision in the Taperloc group and no femoral component revisions in the patients who were octogenarians or who had rheumatoid arthritis. Design features (collarless tapered wedge fit, circumferentially porous-coated) virtually ensure bone ingrowth and are thought to be responsible for the excellent clinical results and longevity.

Comparison of oral fluconazole and itraconazole for progressive, nonmeningeal coccidioidomycosis. A randomized, double-blind trial. Mycoses Study Group.

BACKGROUND: In previous open-label noncomparative clinical trials, both fluconazole and itraconazole were effective therapy for progressive forms of coccidioidomycosis. OBJECTIVE: To determine whether fluconazole or itraconazole is superior for treatment of nonmeningeal progressive coccidioidal infections. DESIGN: Randomized, double-blind, placebo-controlled trial. SETTING: 7 treatment centers in California, Arizona, and Texas. PATIENTS: 198 patients with chronic pulmonary, soft tissue, or skeletal coccidioidal infections. INTERVENTION: Oral fluconazole, 400 mg/d, or itraconazole, 200 mg twice daily. MEASUREMENTS: After 4, 8, and 12 months, a predefined scoring system was used to assess severity of infection. Findings were compared with those at baseline. RESULTS: Overall, 50% of patients (47 of 94) and 63% of patients (61 of 97) responded to 8 months of treatment with fluconazole and itraconazole, respectively (difference, 13 percentage points [95% CI, -2 to 28 percentage points]; P = 0.08). Patients with skeletal infections responded twice as frequently to itraconazole as to fluconazole. By 12 months, 57% of patients had responded to fluconazole and 72% had responded to itraconazole (difference, 15 percentage points [CI, 0.003 to 30 percentage points]; P = 0.05). Soft tissue disease was associated with increased likelihood of response, as in previous studies. Azole drug was detected in serum specimens from all but 3 patients; however, drug concentrations were not helpful in predicting outcome. Relapse rates after discontinuation of therapy did not differ significantly between groups (28% after fluconazole treatment and 18% after itraconazole treatment). Both drugs were well tolerated. CONCLUSIONS: Neither fluconazole nor itraconazole showed statistically superior efficacy in nonmeningeal coccidioidomycosis, although there is a trend toward slightly greater efficacy with itraconazole at the doses studied.

Opportunities and challenges in Medicaid managed care: the experience in Maryland.

OBJECTIVE: The effects of the Maryland Medicaid mandatory managed care programs on Medicaid beneficiaries are examined with the main objective of gaining insight into the initial experience and beneficiary satisfaction with Maryland's Medicaid program. The background of the Maryland Medicaid system, initial implementation, results of beneficiary satisfaction surveys, and future concerns are discussed. STUDY DESIGN: An observational study based on survey data. DATA AND METHODS: Beneficiary surveys mailed to adult and child participants in HealthChoice and the Rare and Expensive Case Management (REM) Medicaid programs in Maryland are analyzed. Descriptive univariate and bivariate data statistics are used. RESULTS: The 4 questions rating satisfaction with primary care provider (PCP), relevant specialists, all providers, and the overall health plan indicate high levels of satisfaction in both adult and child populations. CONCLUSIONS: The Maryland Medicaid programs appear to have met the goal of providing a comprehensive, coordinated healthcare system of quality care during their first year of operation. The satisfaction of these beneficiaries suggests that with an appropriate risk-adjusted capitation approach, managed care organizations (MCOs) can successfully provide for even the most complex needs of Medicaid members.

Pseudoaneurysm of femoral artery after revision total hip arthroplasty with a constrained cup.

Delayed vascular injury after revision total hip arthroplasty is a rare and unusual complication. We report a case of a mechanical complication in which migration of a constraining ring locking mechanism used during a revision total hip arthroplasty caused a pseudoaneurysm of the common femoral artery.

The bearing surface in total hip arthroplasty: evolution or revolution.

Highly cross-linked polyethylenes represent a new class of polyethylenes that demonstrate dramatic improvements of wear characteristics in laboratory tests. Cross-linked polyethylenes can be manufactured by a number of methods, all of which lead to some changes in the physical properties of the polyethylene. The very limited clinical information about cross-linked polyethylenes available has been favorable. Cross-linked polyethylenes appear to hold promise as an alternative to conventional polyethylene and to hard-on-hard bearing surfaces, but much more clinical information will be required before they can be recommended for routine clinical use.

Routine follow-up office visits after total joint replacement: do asymptomatic patients wish to comply?

A total of 100 patients presenting for routine office follow-up after total hip or knee arthroplasty completed questionnaires evaluating whether they preferred to come to the office for routine follow-up evaluation or whether they would have preferred an evaluation without an office visit. Of 100 patients, 45 would have preferred not to come into the office for a routine evaluation. They were content to mail completed questionnaires and radiographs to their physicians. The other 55 patients preferred office visits. These 2 groups were comparable for age, sex, height, weight, and number of surgeries (P > .11) Preoperative and postoperative scores were similar between the 2 groups (P > .39). None of the patients that would have preferred not to come in to the office believed that quality of care would be compromised. A significant number (45%) of patients would prefer not to come to the office because of the wages saved and time spared. Routine office visits may be eliminated for these patients through the use of health outcome devices, such as the SF-36, along with routine radiographs. The potential to decrease healthcare costs and increase patient satisfaction warrants the identification of these patients. Assessment of the effect on quality of care with elimination of routine follow-up visits requires further study.

Prospective multicenter surveillance study of funguria in hospitalized patients. The National Institute for Allergy and Infectious Diseases (NIAID) Mycoses Study Group.

Although fungal urinary tract infections are an increasing nosocomial problem, the significance of funguria is still not clear. This multicenter prospective surveillance study of 861 patients was undertaken to define the epidemiology, management, and outcomes of funguria. Diabetes mellitus was present in 39% of patients, urinary tract abnormalities in 37.7%, and malignancy in 22.2%; only 10.9% had no underlying illnesses. Concomitant nonfungal infections were present in 85%, 90% had received antimicrobial agents, and 83.2% had urinary tract drainage devices. Candida albicans was found in 51.8% of patients and Candida glabrata in 15.6%. Microbiological and clinical outcomes were documented for 530 (61.6%) of the 861 patients. No specific therapy for funguria was given to 155 patients, and the yeast cleared from the urine of 117 (75.5%) of them. Of the 116 patients who had a catheter removed as the only treatment, the funguria cleared in 41 (35.3%). Antifungal therapy was given to 259 patients, eradicating funguria in 130 (50.2%). The rate of eradication with fluconazole was 45.5%, and with amphotericin B bladder irrigation it was 54.4%. Only 7 patients (1.3%) had documented candidemia. The mortality rate was 19.8%, reflecting the multiple serious underlying illnesses found in these patients with funguria.

Continuity of care: is it cost effective?

OBJECTIVE: To examine the association between the degree of healthcare provider continuity and healthcare utilization and costs. STUDY DESIGN: A longitudinal, prospective, observational study. PATIENTS AND METHODS: Data on patients with arthritis, asthma, epigastric pain/peptic ulcer disease, hypertension, and otitis media were collected at each of 6 health maintenance organizations (HMOs). Outcome variables included the number of prescriptions for the target disease and the cost, total number of prescriptions and the cost, the number of outpatient visits, and the number of hospital admissions. Disease-specific severity of illness, type of visit, and provider information were obtained at each encounter. HMO profit status, visit copay, gatekeeper strictness, formulary limitations, use of multisource (generic) drugs, gender, number of months in the study, age, and severity of illness were controlled in the analyses. RESULTS: There were 12,997 patients followed for more than 99,000 outpatient visits, 1000 hospitalizations, and more than 240,000 prescriptions. Increasing the number of primary or specialty care providers a patient encountered during the study generally was associated with increased utilization and costs when HMO and patient characteristics were controlled. The number of specialty care providers also increased as the number of primary care providers increased. The incremental increase in pharmacy costs per patient per year with each additional provider ranged between $19 in subjects with otitis media to $58 in subjects with hypertension. CONCLUSIONS: Continuity of care was associated with a reduction in resource utilization and costs. As healthcare delivery systems are designed, care continuity should be promoted.

Efficacy of irrigation for removal of particulate debris after cemented total knee arthroplasty.

We studied the amount of particulate debris removed with pulsatile lavage irrigation before and after component implantation in 13 consecutive patients undergoing primary cemented total knee arthroplasty (TKA) done by a single surgeon. Before component implantation, the knees were copiously lavaged with 3 L of pressurized irrigant; all fluid was collected in 1 aliquot using standard wall suction canisters. After cementing the components in place, another 3 L of pressurized irrigant was used; this fluid was collected in 3 sequentially labeled 1-L aliquots. Collected fluids were centrifuged, and the residue was washed, recentrifuged, and dried. Residual particulate debris was quantitated by weight. An average of 537 mg/L (range, 16-1,406 mg/L) of debris were removed before implantation with 3 L of irrigation. An average of 217 mg/L (range, 31-999 mg/L), 52 mg/L (range, 0-189 mg/L), and 49 mg/L (range, 1-185 mg/L) of debris was removed after implantation with each of the additional liters. Using analysis of variance testing, there was a statistically significant difference between the amount of debris removed with 3 L and after 4 L (P = .02) and 5 L (P = .03) of irrigant. There was no statistical difference between irrigation with 5 L and 6 L of irrigation (P = .92). The residua particulate debris was also analyzed to determine the relative amounts of bone-soluable organics and polymethyl methacrylate (PMMA). Before implantation, the residual debris, by weight, consisted of 79% bone and 21% soluable organics. We found on average that after implantation of components the specimens contained 53% bone and 47% PMMA and soluable organics by weight. We believe that despite careful implantation and meticulous cement technique, large amounts of debris, including bone and PMMA, remain after TKA, which require at least several liters of pulsatile lavage to remove. Removal of this particulate debris may decrease third-body polyethylene wear.