Comparing physician-specific two-year patient outcomes after coronary angiography: methodologic issues and results.

OBJECTIVES: We sought to evaluate methodologies to compare physician-related long-term patient outcomes appropriately. BACKGROUND: Evaluation of physicians on the basis of short-term patient outcome is becoming widely practiced. These analyses fail to consider the importance of long-term outcome, and methods appropriate to such an analysis are poorly defined. METHODS: All patients undergoing coronary angiography between 1992 and 1994 who received all of their cardiac care at our institution were followed for 27+/-13 months (mean+/-SD). Patients (n = 754) were cared for by one or more of 17 staff physicians. Risk-adjusted models were developed for four candidate clinical end points and cost. Physicians were then evaluated for each outcome measure. RESULTS: Of the clinical end points, death could be modeled most accurately (c-statistic = 0.83). The c-statistics for other end points ranged from 0.63 to 0.70. Physicians with outcomes statistically different (p < 0.05) from other physicians were identified more commonly than would be expected from the play of chance (p = 0.005). However, improvement in the c-statistics by the addition of physician identifiers was very modest. Physician's evaluations by the four measures of clinical outcome were variably correlated (r = .00 to .85). Graphic display of clinical and cost results for each physician did identify certain physicians who might be judged to provide more cost-effective care than others. CONCLUSIONS: Although comparisons of groups of physicians on the basis of long-term patient outcomes may have merit, individual physician-to-physician comparisons will be more difficult, owing to 1) multiple physicians contributing care to individual patients; 2) the poor predictive capacity of models other than that for survival; and 3) the modest apparent impact of differences in physician providers on long-term patient outcome. With these caveats in mind, modeling to compare patient outcomes of individual physicians with homogeneous patient populations or to identify gross outliers (good or bad) may be practicable in some patient-care systems, but may be inappropriate in others.

Vertebral osteomyelitis presenting as spinal compression fracture. Six patients with underlying osteoporosis.

Six patients with osteoporosis had vertebral osteomyelitis (VO) with infection of a single vertebra that presented with a collapsed vertebral body, thought to be a simple compression fracture. The resulting delay in correctly diagnosing VO was associated with disabling sequelae in a high proportion of cases. This distinctive presentation accounted for 13% of all hospitalized patients with VO and 2.4% of inpatients with osteoporotic compression fractures during the last five years; it may be more common than suggested by the paucity of published cases. In patients with osteoporosis and vertebral compression fractures, osteomyelitis should be considered when there is severe back pain, persistent unexplained fever, unexplained elevation of the erythrocyte sedimentation rate, or bacteremia without an obvious extravertebral focus of infection, particularly if the patient is immunocompromised. Early biopsy and culture of the collapsed vertebral body will facilitate diagnosis and therapy.

Viral hepatitis. A population-based study in Rochester, Minn, 1971-1980.

The incidence of acute viral hepatitis among Rochester, Minn, residents 1971-1980 was 28.6 per 100,000 person-years (p-y) (age- and sex-adjusted to the 1980 white population in the United States). The adjusted incidence of hepatitis B (12.9 per 100,000 p-y) was somewhat less than for hepatitis non-B (15.6 per 100,000 p-y). Each type was more frequent among young adults, especially males. The incidence of hepatitis was greater among those employed in the health service industry than among nonmedical employees (53.4 vs 20.0 per 100,000 p-y). Medical employees had nearly a fivefold increased incidence of hepatitis B and a twofold increased incidence of hepatitis non-B. Exposure to known hepatitis cases was common, but other possible causative factors were not frequent. In this midwestern community, the incidence of acute viral hepatitis is substantial, with medical employees at significantly increased risk.

Nosocomial clonal dissemination of methicillin-resistant Staphylococcus aureus. Elucidation by plasmid analysis.

An outbreak of methicillin sodium-resistant Staphylococcus aureus (MRSA) infection and colonization, mainly centered in the vascular surgery service, occurred in a 1000-bed tertiary care center between December 1983 and December 1984. Methicillin-resistant S aureus isolated before and during the outbreak was studied by both bacteriophage typing and by restriction endonuclease digestion of bacterial plasmid DNA. Bacteriophage typing was discrepant in nine (56%) of the 17 repeated analyses compared with one (3.4%) of the 29 for plasmid profiling. These typing methods revealed that the epidemic strain was introduced to the hospital from the community 15 months before the outbreak. The outbreak was caused by cross-transmission of the epidemic strain by health care personnel and was controlled by treatment of colonized personnel, education of personnel, and institution of barrier precautions for colonized or infected patients. Plasmid profiling with restriction endonuclease digestion was easier, more rapid, and more specific than bacteriophage typing in the evaluation of this outbreak.

Fungal prosthetic valve endocarditis in 16 patients. An 11-year experience in a tertiary care hospital.

Fungal prosthetic valve endocarditis (PVE) is an infrequent but serious complication of valve replacement surgery. To examine its long-term outcome, we retrospectively studied 16 patients with 19 episodes of definite fungal PVE. The mean age was 51 years (range, 27-71 yr). Onset of fungal PVE ranged from 8 days to 3.4 years after valve replacement. Candida albicans was the most common (56%) pathogen isolated. A portal of entry was identified in only 25% of the patients; the presence of intravascular catheters (50%) and prior bacterial endocarditis (38%) were leading predisposing factors. Fever (83%) was the most consistent clinical finding. Potentially serious embolic events, particularly strokes (32%), were common at presentation. Transesophageal echocardiography (sensitivity = 100%) was more useful than transthoracic echocardiography (sensitivity = 60%) in detecting lesions due to fungal PVE. Combined valve replacement surgery and amphotericin B (mean total dose of 1.8 g) in 15 patients resulted in an 87% in-hospital survival and 67% overall survival with a mean follow-up of 4.5 years (range, 5 mo to 16 yr). Two patients had 3 late relapses of fungal PVE up to 9 years after the preceding episode. Each relapse was treated with repeat valve replacement and amphotericin B; in addition, oral azole was utilized for chronic suppression, although the efficacy of this strategy remains unproven. Because of the possibility of relapse, long-term follow-up is essential even after surgical and prolonged antifungal therapy.

Left-sided endocarditis due to Pseudomonas aeruginosa. A report of 10 cases and review of the literature.

Ten confirmed cases of left-sided endocarditis due to Pseudomonas aeruginosa were reported in detail and the English literature was reviewed. In recent years, venous access (usually illicit) has been the major predisposing factor to this infection and abuse of pentazocine and tripelennamine has been particularly associated with endocarditis due to this organism. This infection involves previously damaged as well as normal valves. The development of congestive heart failure did not adversely affect the prognosis of this infection. However, the development of azotemia was associated with a greater likelihood of a fatal outcome. In the current series, deaths were due to uncontrolled infection. This often occurred despite inhibitory and bactericidal activity in serum generally considered adequate for treatment of endocarditis. Medical treatment alone rarely produced cure of infection. Our experience with a high frequency of major vessel embolization (4/10) and the improved survival after medical/surgical treatment suggests that prompt valve replacement combined with high doses of an aminoglycoside plus carbenicillin or ticarcillin provide the best opportunity for successful outcome in patients with left-sided endocarditis due to P. aeruginosa.

Neurologic complications of endocarditis: a 12-year experience.

We reviewed the neurologic complications in 113 patients with native and 62 patients with prosthetic valve endocarditis. Neurologic complications occurred with the same frequency (35.3% vs 38.7%) and distribution among the two groups. Death occurred in 20.6% of patients with neurologic complications and in 13.6% of patients without neurologic complications (p = 0.23). Staphylococcus aureus endocarditis correlated statistically with the development of neurologic complications (p less than 0.01) and death (p less than 0.01). Among 50 patients discharged from the hospital after receiving only medical treatment for native valve endocarditis, and followed for a mean period of 48 months, there was one patient with mitral valve prolapse and stroke. We conclude that (1) neurologic complications occur with the same frequency in native and prosthetic valve endocarditis, (2) S aureus endocarditis increases the risk of neurologic complications and death, (3) mortality is not significantly increased in patients with neurologic complications, and (4) an episode of treated native valve endocarditis does not increase the natural history of stroke in valvular disease.

Central nervous system histoplasmosis with obstructive hydrocephalus.

A case of central nervous system histoplasmosis complicated by obstruction of the fourth ventricle is described. The patient rarely exhibited systemic symtoms of infection despite positive cultures for Histoplasma capsulatum from bone marrow, blood and urine. Infection recurred despite the administration of a total of 5 g of systemic amphotericin B. An additional course combined with intrathecal amphotericin B was terminated because of transverse myelitis.

Gallium scanning for the detection of abdominal abscesses.

The use of gallium scanning for the diagnosis of abdominal abscesses was studied in 59 cases in which the diagnosis was verified by laparotomy or autopsy. In 23 of the cases the patients were subsequently found to have abdominal abscesses and 11 patients had abdominal malignancies. Excluding the latter group, the scan was specific for abscess in 86 per cent of the cases and sensitive for abscess in 67 per cent. Recent abdominal incisions, concurrent intestinal, hepatic, biliary and pancreatic inflammatory conditions, without abscess, seldom caused gallium accumulation. False-negative scans (18 per cent) occurred more frequently than false-positive scans (8 per cent) and were seen in patients with large (often palpable) masses of short evolution and with secondarily infected lesions such as hematomas and pseudocysts. The gallium scan was of little help in the search for abdominal lesions in patients with obscure febrile illnesses without abdominal symptoms or signs.

Infections of the cardiac suture line after left ventricular surgery.

PURPOSE: Infections of the cardiac suture line after left ventricular surgery are rare but may be fatal if not diagnosed promptly and treated effectively. In order to alert physicians to this entity, we reviewed data from three patients who presented at the Cleveland Clinic Hospital and from 22 patients in the literature. PATIENTS AND METHODS: The three patients in the current report underwent treatment at the Cleveland Clinic Hospital. Additional cases of infection of the left ventricular suture line were identified by reviewing the English literature pertaining to surgery for left ventricular aneurysms, pseudoaneurysms, and postoperative cardiac infections. RESULTS: Infection presented on average 16 months after surgery with cardiocutaneous fistulae, chest wall masses, hemoptysis or other pleuropulmonary symptoms, or systemic illness with bacteremia resembling endocarditis. Staphylococci and gram-negative bacilli were the most frequent pathogens. Diagnosis was often delayed and mortality was high. Left ventricular false aneurysms were identified in 15 of the 25 patients. Bleeding from sinuses in the chest wall or epigastrium or repeated hemoptysis were important clinical clues. In some instances, ill-advised surgical or instrumental procedures precipitated life-threatening hemorrhage. Treatment with antibiotics alone was insufficient. Excision of all infected sutures and Teflon pledgets and adequate debridement of the infected suture line were required to achieve cures. CONCLUSION: Since infection of the left ventricular suture line has protean clinical manifestations and may present months or years after the initial surgery, a high index of suspicion is of paramount importance in diagnosing the condition. Institution of cardiopulmonary bypass and reoperation through median sternotomy is recommended to achieve a cure.

A multivariate analysis of risk factors for pneumonia following cardiac transplantation.

Fifty cardiac transplant recipients were followed over a 34-month period for evidence of pneumonia that developed in twelve patients. Potential risk factors evaluated fell into three categories: demographic (age, sex, race, and underlying cardiac disease); pretransplant status (hospitalized, intubated, pulmonary infiltrate, requirement for antibiotics, or the need for a ventricular assist device); and posttransplant therapy (amount and type of blood products, prolonged endotracheal intubation or reintubation, use of ventricular assist devices, immunosuppressive protocols, lymphocyte subset ratios, and occurrence of rejection, leukopenia, or CMV infection). The Cox proportional hazards model identified posttransplant reintubation (P = 0.009) and the use of protocols employing larger steroid dosages (P = 0.02) as significant risk factors for pneumonia. In a separate analysis, the occurrence of pneumonia was shown to be a significant risk factor for mortality (P = 0.018).

Legionnaires' disease: a review of the epidemiology and clinical manifestations of a newly recognized infection.

During the 2-year period 1977 through 1979, 26 patients with Legionnaires' disease were seen at the Mayo Clinic and affiliated hospitals. The patients ranged in age from 17 to 81 years with a median of 51 years. Twelve (46%) were immunologically compromised. Most of the other patients had underlying chronic tobacco bronchitis. Hectic fever, cough, and diarrhea were common symptoms. Chest radiographs showed patchy perihilar infiltrates that often progressed to consolidation. Diagnosis was made by indirect fluorescent antibody testing in 15 patients (58%), but in no case was the test diagnostic during the first week of illness. In seven patients the diagnosis was established by positive direct flourescent antibody testing of lung tissue, in two cases by culture of lung tissue, and in one case each by direct fluorescent antibody positivity of sputum or bronchial washing. Of the 26 patients, 3 (12%) required hemodialysis for acute renal failure and 5 (19%) died. A favorable clinical response to therapy with erythromycin was noted. The differential diagnosis of Legionnaires' disease must include other bacterial pneumonias, as well as mycoplasma, psittacosis, Q fever, and viral pneumonia. For critically ill patients, open-lung biopsy may be necessary to provide a rapid diagnosis. Current evidence suggests that erythromycin alone or in combination with rifampin is the treatment of choice. A 3-week course of therapy is recommended in order to prevent relapse.

Antimicrobials commonly used for urinary tract infection: sulfonamides, trimethoprim-sulfamethoxazole, nitrofurantoin, nalidixic acid.

Sulfonamides, trimethoprim-sulfamethoxazole, nitrofurantoin, and nalidixic acid are excreted in the urine in high concentration and, with the exception of Pseudomonas aeruginosa and Serratia marcescens, are all active in vitro against usual aerobic gram-negative bacteria. They are useful for treatment of uncomplicated lower urinary tract infections in an outpatient setting but are probably not indicated for acute upper tract infections, especially in the hospitalized patient. Trimethoprim-sulfamethoxazole is a very effective combination agent in vitro, has appealing pharmacokinetic properties, and is usually well tolerated by patients. However, it is more expensive than the sulfonamides and is ordinarily not indicated for initial treatment. A sulfonamide is in our view still the agent of first choice, with either nitrofurantoin or nalidixic acid as an alternative if sulfonamides cannot be tolerated.

A sporadic case of pneumonia due to legionnaires disease.

A recently recognized sporadic case of "legionnaires' disease" occuring in Minnesota is described. The patient, a previously healthy 58-year-old woman, became seriously ill with progressive pneumonia in the spring of 1977. Her illness was unresponsive to the usual antibiotics (penicillin and cephalothin) prescribed for community-acquired bacterial pneumonia. Treatment with doxycycline appeared to be effective. Diagnosis was made by special serologic studies performed at the Center for Disease Control, Atlanta. As reported in the epidemic form of the disease, no secondary cases were noted in family, associates, or hospital workers exposed to the patient during her acute illness. This disease may be more common than has previously been recognized, for serologic test for the unusual bacterium have only recently become available.

Gentamicin-resistant Pseudomonas aeruginosa: Mayo Clinic Experience, 1970-1976.

During the period 1970 through 1976, there were 144 patients from whom gentamicin-resistant Pseudomonas aeruginosa (minimum inhibitory concentration [MIC], more than 5 microgram/ml) was isolated. In 20(21 percent) of the 95 patients who acquired such organisms within our institutions, the occurrence was considered clinically significant. Factors that favored the appearance of gentamicin-resistant P. aeruginosa included prolonged hospitalization, previous antibiotic treatment, increased gentamicin usage, underlying disease, and instrumentation (70 percent). Virulence of gentamicin-resistant isolates appeared less than that of susceptible organisms, with bacteremia due to these isolates occurring in only three cases. Resistant isolates with MICs for gentamicin of 8 to 16 microgram/ml were more susceptible to tobramycin than to amikacin, whereas isolates with MICs for gentamicin of 64 microgram/ml or greater were more susceptible to amikacin than to tobramycin. Eighty percent of all strains were susceptible to 128 microgram/ml or less of carvenicillin. Favorable results occurred in 12 or 13 cases treated with gentamicin plus carbenicillin, whereas treatment with either of these agents alone resulted in failure or relapse in 7 of 14 cases.

Treatment of primary pulmonary blastomycosis: results of long-term follow-up.

Twenty-three patients with pulmonary blastomycosis were seen from 1950 to 1975, and long-term follow-up, ranging from 2 to 21 years, was available for 19 patients. Four major treatment groups were identified: surgery alone (eight patients), amphotericin B alone (four patients), surgery and amphotericin B (three patients), and observation (four patients). Six of eight patients who had surgical resection alone were cured; one patient died of cardiopulmonary complications in the immediate postoperative period, and one had a relapse 15 years later. The remaining 11 patients in the three other treatment groups were cured. Although systemic therapy with amphotericin B is clearly indicated for progressive, disseminated, or incompletely resected disease, a period of observation is indicated for nonprogressive or totally resected pulmonary blastomycosis.

Antifungal agents used for deep-seated mycotic infections.

The main antifungal agents used for deep-seated mycotic infections are the broad-spectrum antifungal drug amphotericin B, the narrow-spectrum agent flucytosine, and the newer broad-spectrum agents miconazole and ketoconazole. Amphotericin B remains the cornerstone of antifungal therapy. For the treatment of cryptococcal meningitis, the current recommendation is for the combined use of amphotericin B and flucytosine. 2-Hydroxystilbamidine is used only in indolent cases of blastomycosis; however, this condition is usually treated with amphotericin B. Clinical experience with the newer agents is limited. Not all patients from whom fungal agents have been isolated require treatment; the extent of the fungal infection should be determined, when possible, for evaluation of the need for treatment.

Antimicrobial agents in urinary tract infections.

Effective antimicrobial therapy for most urinary tract infections has been available since the sulfonamide era. Innovations in chemotherapy now include single-dose treatment of acute bacterial cystitis and acute urethral syndrome and effective suppression of recurrent bacteriuria with low-dose antimicrobial agents.

The aminoglycosides. Streptomycin, kanamycin, gentamicin, tobramycin, amikacin, netilmicin, sisomicin.

Despite their toxicity, the aminoglycosides remain useful and are often the first choice in the treatment of serious infections due to gram-negative bacilli. Nephrotoxicity has restricted the indications for neomycin to topical and oral use. Emergence of resistant organisms has limited the use of streptomycin to a few specific conditions. Gentamicin, tobramycin, and amikacin are effective against a broad spectrum of gram-negative bacilli including Pseudomonas aeruginosa. Amikacin is the aminoglycoside of choice when gentamicin resistance is prevalent.

Acute cystitis: a prospective study of laboratory tests and duration of therapy.

The efficacy of single-dose therapy with trimethoprim-sulfamethoxazole (TMP-SMZ) and the cost-effectiveness of routine urinalyses and cultures were studied in a prospective randomized trial of 200 women who presented with symptoms of acute lower urinary tract infection. Without the physician's knowledge of the results of urinalysis or culture, the patients were randomly assigned to receive either a single dose or a 10-day multiple-dose course of TMP-SMZ and were followed up for 6 months. Of the 136 patients with positive urine cultures, 68 received single-dose therapy with TMP-SMZ--10 of whom had relapses--and 68 received multiple-dose therapy with TMP-SMZ--only 2 of whom had relapses (P less than 0.02). Fifteen patients in each treatment group experienced reinfection. Side effects of rash and vaginitis were more common in patients who received multiple-dose therapy, but they were mild and well tolerated. Of the 51 patients with urethral syndrome, 48 became asymptomatic after therapy. None of the following tests predicted treatment outcome: pretreatment urinalysis, urine culture or susceptibility testing, antibody-coated bacteria testing, or routine follow-up urinalyses or urine cultures. Empiric therapy with TMP-SMZ in selected women with symptoms of acute uncomplicated urinary tract infection seems practical, safe, and cost-efficient. Considerable savings can be achieved by reserving urinalyses and urine cultures for patients with persistent or recurrent symptoms. Higher cure rates can be expected in patients who receive a standard 10-day course of therapy with TMP-SMZ compared with those who receive single-dose therapy with TMP-SMZ.