Corinne Dulberg

Association of cisplatin nephrotoxicity with patient characteristics and cisplatin administration methods

Objective: To assess factors that affect cisplatin nephrotoxicity. Methods: In 425 patients treated with cisplatin, we assessed the effect of pretreatment factors and treatment conditions on the rise in serum creatinine with the first course of cisplatin, on the maximum rise in serum creatinine over the entire course of the cisplatin therapy, and on residual nephrotoxicity after the last cisplatin treatment ended. (Because of the nature of the relationship between serum creatinine and creatinine clearance, rise in serum creatinine was divided by pretreatment creatinine squared.) Patients were dichotomized into the upper quartile versus the lower three quartiles of degree of nephrotoxicity. Multivariate analyses were based on logistic regression, controlling for cisplatin dose per course. Results: Controlling for cisplatin dose per course, factors most closely associated with nephrotoxicity during the first course of cisplatin were: serum albumin and potassium, body surface area, and administration of cisplatin over 2–5 days per course vs 1 day (negative associations). Controlling for cisplatin dose per course, the single factor most closely associated with maximum life-time cisplatin nephrotoxicity was concurrent use of a vinca alkaloid (negative association). Controlling for cisplatin dose per course, factors most closely associated with residual nephrotoxicity after the end of cisplatin therapy were cumulative dose of cisplatin, concurrent use of metoclopramide (positive associations), uric acid and concurrent use of phenytoin and a vinca alkaloid (negative associations). The association of nephrotoxicity with uric acid and with body surface area was felt to be an artifact resulting from its positive association with pretreatment serum creatinine. Nephrotoxicity during the first course of cisplatin also correlated significantly with autopsy kidney cortex platinum concentrations in 77 evaluable patients. Conclusions: (1) While several factors correlated with cisplatin nephrotoxicity, most of the observed nephrotoxicity was not explained by the variables identified. (2) While most patients received intravenous hydration, patients receiving high hydration volumes did not have significantly less nephrotoxicity than patients receiving lower hydration volumes. (3) Of the variables identified, serum albumin, metoclopramide and phenytoin may have affected nephrotoxicity by altering cisplatin uptake into or distribution within the kidney.

A prospective clinical audit of neonatal resuscitation practices in Canada.

PurposeThis is a prospective audit to determine the frequency of resuscitation interventions in the clinical setting and to compare self-reports of clinical performance with the existing Neonatal Resuscitation Program (NRP) and Canadian National Guidelines for Neonatal Resuscitation. SUBJECTSFifty-six level I, II, and III hospitals in Canada participated. Any infant requiring resuscitation, as defined by the need for at least positive pressure ventilation (PPV), was eligible for inclusion (n = 783 resuscitations). DESIGN AND METHODSA prospective self-report audit was chosen and data were collected over a 6-month period in 1998. The audit focused on the use of PPV, intubation, chest compressions, free-flow oxygen, or medications during the resuscitation. The infants temperature at the end of resuscitation was also noted. The data were analyzed with descriptive statistics. The composition of the resuscitation team and their NRP certification status were recorded. PRINCIPAL RESULTSThe need for resuscitation was not anticipated in 76% of the cases (596 of 783). Errors in the sequencing of care, such as delays in initiating PPV, provision of chest compressions before or without establishing an airway and ventilatory support, and administering naloxone before PPV, were reported. Resuscitations attended by a team of NRP certified providers had improved sequencing when compared with those in which only some individual providers were certified. Chest compressions were provided in 8% of the cases (65 of 783). Medications were used in 14% (113/783) of all cases. Providers in level I hospitals performed chest compressions more frequently than those in level II and III settings. At the end of the resuscitation, 27% of the infants were hypothermic (142 of 520), and 25% were hyperthermic (128 of 520). Overall, 52% were out of the normal neutral range. CONCLUSIONSClear differences between the NRP guidelines and actual clinical practice were shown. A high rate of unanticipated resuscitations, delivery room medications, and chest compressions was described. Postresuscitation hypothermia or hyperthermia were common. Improved sequencing was noted when the entire resuscitation team was NRP certified. Certification in NRP does not assure competency, nor does it ensure compliance with established standards of care.

A randomized, controlled trial of radiograph ordering for extremity trauma in a pediatric emergency department

STUDY OBJECTIVES The objectives of this study were to determine whether triage nurses using the Brand protocol would order fewer radiographs than would physicians carrying out standard practice procedures, without missing an increased number of joint or bone injuries; the test characteristics and the interobserver reliability of the Brand protocol; and whether having triage nurses order radiographs could reduce total patient waiting time in the emergency department. DESIGN Randomized, controlled trial. SETTING The ED of a free-standing childrens hospital with approximately 55,000 visits annually. TYPE OF PARTICIPANTS Children less than 18 years of age who had a history of extremity trauma in the preceding seven days. INTERVENTIONS Triage nurses applied the Brand protocol to determine the need for a radiograph. MEASUREMENTS AND RESULTS Of the Brand protocol group, 81.9% had radiographs ordered compared with 87.1% of the control group (P = .03). The percent of positive radiographs was 40.8% in the Brand protocol group compared with 42.6% in the control group (P = .21). There were 3.2% (16) missed radiographic findings in the Brand protocol group compared with none in the control group (P < .001). Patients randomized to the Brand protocol group spent 3.3 hours in the ED compared with 3.6 hours for the control group (P < .001). CONCLUSION Having triage nurses use the Brand protocol reduced the number of radiographs ordered but at the same time increased the number of missed radiographic findings. However, having triage nurses order radiographs also significantly shortened waiting time in the ED.

Children's Memory for Locations Visited: Importance of Movement and Choice

In 2 experiments, the effect of active or passive mobility and active or passive choice experiences on childrens memory for locations visited while retrieving puzzle pieces hidden in a large room were examined. In the first experiment, fifty-two 6- and 7-year-old children were randomly assigned to 1 of 4 training conditions: active choice-active movement, active choice-passive movement, passive choice-active movement, and passive choice-passive movement. After 3 training trials, all children were tested in the active choice-active movement condition. A 2 (movement) x 2 (choice) factorial multivariate analysis of variance revealed a significant main effect for movement. Children who had moved actively to find the puzzle pieces in the training trials had (a) a greater percentage correct, (b) more correct visits between errors, and (c) a later visit of first error than children who had been passively pushed in a wheelchair. Making active choices in training did not significantly affect performance. A second experiment used identical procedures but tested 32 children with mobility difficulties who regularly used a wheelchair because of either cerebral palsy or spina bifida. Children from this group who moved actively during training to retrieve the puzzle pieces also performed better on testing. Implications of the results for children with disabilities and for developmental theory are discussed.

Factors affecting human autopsy kidney-cortex and kidney-medulla platinum concentrations after cisplatin administration

The objective of this study was to determine factors that affect cisplatin concentrations in human kidney cortex. We used flameless atomic absorption spectrophotometry to assay platinum in autopsy specimens of kidney cortex obtained from 83 cisplatin-treated patients. Concentrations were correlated with pretreatment factors and treatment conditions using univariate nonparametric statistics. Hierarchical stepwise multiple regression analyses of transformed (to normalize) data were then used to assess which factors were most important, controlling for other factors. Kidney-cortex platinum concentrations varied from 0 to 14.8 μg/g (median, 2.04 μg/g). The cumulative lifetime dose of cisplatin ranged from 10 to 1120 mg/m2 (median, 112 mg/m2). The time from the last cisplatin dose to death was <1–609 days (median, 38 days). According to univariate statistics, factors that correlated (P<0.05) with kidney-cortex platinum concentrations were the cisplatin dose per course, the pretreatment serum urea level, metoclopramide use (positive correlations), the time from the last cisplatin treatment to death, and the pretreatment serum albumin value (negative correlations). Factors that approached significance (0.05≤P≤0.10) were a history of hypertension, hyperbilirubinemia (positive), the serum calcium level, and phenytoin use (negative). In the multiple regression analysis, after controlling for the cisplatin dose per course and the time from the last treatment to death, only concurrent metoclopramide and phenytoin use entered the model. The hydration volume did not affect corrected kidney-cortex or kidney-medulla platinum concentrations. The following conclusions were reached: (1) it may be feasible to use lower hydration volumes than those used routinely, (2) any effect of hydration volume on cisplatin nephrotoxicity may not be mediated via a reduction in kidney-cortex platinum concentrations, (3) higher cisplatin doses might be tolerated with new 5-hydroxytryptamine-3 (5HT-3) antiemetics than were tolerated with metoclopramide, and (4) phenytoin should be tested for its ability to reduce cisplatin nephrotoxicity.

Radiograph ordering: Agreement between the triage nurse and the physician in a pediatric emergency department

We hypothesized that the triage nurse in a busy pediatric emergency department (ED) could accurately order radiographs, ultimately reducing patient waiting time. Protocols utilized to reduce patient waiting time are of importance in busy emergency departments. All patients registering at the ED of the Childrens Hospital in a one-week period were entered into the study. The triage nurse documented whether they would send the patient for a radiograph, if allowed to, and were asked to designate a specific radiograph. The radiographs that were actually ordered by the physicians were subsequently obtained from radiology records. Data were analyzed for agreement beyond chance (kappa), positive (PPV) and negative predictive values. The results showed excellent agreement for extremity radiographs and poor agreement for nonextremity radiographs.

Radiograph ordering: Agreement between the Triage Nurse and Physician in a Pediatric Emergency Department

We hypothesized that the triage nurse in a busy pediatric emergency department (ED) could accurately order radiographs, ultimately reducing patient waiting time. Protocols utilized to reduce patient waiting time are of importance in busy emergency departments. All patients registering at the ED of the Childrens Hospital in a one-week period were entered into the study. The triage nurse documented whether they would send the patient for a radiograph, if allowed to, and were asked to designate a specific radiograph. The radiographs that were actually ordered by the physicians were subsequently obtained from radiology records. Data were analyzed for agreement beyond chance (kappa), positive (PPV) and negative predictive values. The results showed excellent agreement for extremity radiographs and poor agreement for nonextremity radiographs.

Population Attributable Risk for Singleton Preterm Birth

Abstract Objective: To determine the prevalence and population attributable risk PAR of sociodemographic and lifestyle risk factors for delivery of premature babies among women giving birth to a single baby and living in Ottawa-Carleton, Canada. Methods: A population-based survey was conducted with a self-administered questionnaire which was completed by women in the immediate postpartum period in hospital. Women who did not complete the questionnaire at that time were followed-up by mail and phone. Women who refused the long questionnaire were asked to complete a short version. Logistic regression analysis was used to identify factors associated with Preterm birth. Adjusted PARs were calculated to examine the contribution of individual risk factors controlling for other selected factors. Results: The full questionnaire was completed by 7940 women (81.7%) and 512 completed a shorter version (5.3%). Key risk factors, identified via logistic regression, were then used to calculate population attributable risk percent. The factors contributing most to preterm birth as calculated by the population attributable risk percent when adjusting for other factors were: primiparity (37.0%), presence of a serious health problem (14.2%), high perceived stress during pregnancy (14.0%), previous preterm birth (13.1%), smoking after the fourth month (9%) short maternal height (6.6%), and previous abortion (5.5%). Conclusion: Understanding the relative contribution of possible modifiable and non-modifiable factors that are associated with preterm birth in a specific population is an important step in the design of prevention strategies. However, preterm prevention programs should not rely on lifestyle modification alone, as these factors make only a modest contribution to the overall population attributable risk for preterm birth.