Outcomes Research Can Lead to More Cost-Efficient Care

Ramesh Sachdeva, MD, PhD, MBA, Center for Outcomes Research and Quality Management, Children's Hospital of Wisconsin;
Kathy Mussatto, RN, BSN, Cardiovascular Surgery, Children's Hospital of Wisconsin

Health care has been a major focus of public and political debate recently, including front-page coverage by The New York Times. Two fundamental questions have emerged.

• How can universal health care be provided to all Americans?
• How can health care be more cost-efficient?

Cost-efficient does not imply cheaper care, but focuses on maximizing the benefit or outcome provided by health care. This includes two concepts - 1) Choosing lower-cost medical care if it provides the same outcome as higher cost medical care, 2) Choosing higher-cost medical care if it provides a better outcome than lower-cost medical care. Since cost-efficient care can be associated with higher and lower costs based on changes in outcome, health care organizations need to link cost with outcomes.

Although outcomes research was developed in the 1960s, this concept only recently gained popularity and support in health care organizations. Children's Hospital of Wisconsin established the Center for Outcomes Research and Quality Management in July 1999. It is one of a few such centers in the country. Joint Commission on Accreditation of Health Care Organizations (JCAHO) also has adopted an outcomes-based evaluation process.

Since 1999, the Center for Outcomes Research has worked with the Cardiovascular Surgery Department at Children's Hospital to facilitate outcomes-based projects. The pediatric CV surgery program at Children's Hospital is one of the leading programs in the country. Over the years, the program has grown tremendously, with more complex surgeries being performed with good outcomes on increasingly sicker children. Literature shows that although comparisons of survival/mortality and length of hospitalization were used as outcome measures, it now is important to measure more subtle outcomes such as functionality and the quality of children's lives after heart surgery. To ensure meaningful comparisons, we use new techniques to adjust the level of risk, such as the Pediatric Risk of Mortality (PRISM) score, as well as health utility measurements to quantify the quality of life.

Several CV outcome research projects are under way including an evaluation of functionality and health-related quality of life in all patients who have received homograft tissue valves. This study involves a phone interview with valve recipients and/or their parents. A critical pathway project also is in progress that involves both critical care and CV surgery. Data is being collected on all patients who have undergone the bidirectional cavo-pulmonary (Glenn) shunt procedure since 1994. A factor analysis of the clinical data will help identify those care practices that contributed to desired outcomes. From this analysis, a critical pathway can be developed. It is anticipated this study will serve as a model for critical pathway development for other procedures.

Another study is designed to compare outcomes in patients undergoing the Norwood procedure for hypoplastic left heart syndrome with patients who have had surgical repair of a congenital diaphragmatic hernia. The study will address variables including health care costs, length of stay and parent-reported quality of life in the two groups. Thoracic transplant outcomes also are being reviewed.

The design and maintenance of prospective databases are very important to future research to provide a storehouse of valuable clinical data for ongoing analysis and outcome monitoring. All of these projects allow us to contribute to the science of caring for children with congenital heart disease.

Acute Renal Failure in the Postoperative Cardiac Patient

Cynthia G. Pan, MD, program director, Nephrology, Children's Hospital of Wisconsin, associate professor, Pediatrics, Medical College of Wisconsin

Acute renal failure following cardiac surgery poses a significant challenge to caregivers. The incidence of acute renal failure in the pediatric population ranges from 2.4 percent to 8 percent and has not changed much in the last decade, despite advances in surgical techniques and intensive care. All causes of acute renal failure in pediatric patients increase mortality, but not surprising, mortality rate more than doubles when renal failure occurs with cardiac surgery.

Pathophysiology
Events leading to postoperative renal dysfunction usually are multifactorial, but diminished renal perfusion secondary to poor myocardial function and hypotension are felt to be the leading factors. Using a variety of techniques, changes in renal blood flow and intrarenal vascular resistance have been demonstrated in patients who develop acute renal failure. In one study using noninvasive techniques of Doppler ultrasound, renal blood flow was assessed in pediatric patients before and after cardiopulmonary bypass surgery for complex congenital defects. They observed that 6.1 percent of patients developed acute renal failure requiring dialysis, and nearly all of affected patients demonstrated reversed diastolic renal blood flow. Reappearance of diastolic blood flow predicted recovery of renal function. Using a measurement of pusatility index (PI) (where PI equals peak systolic velocity minus minimum diastolic velocity divided by the mean velocity of the whole profile), they were able to correlate blood flow patterns with oliguria as well as systolic blood pressure. In patients who had an ASD repair without any complications, no significant changes in PI were demonstrated. The authors suggested the influence of cardiopulmonary bypass itself was not implicated in the changes in renal blood flow. The influence of time on CPB could not be ruled out in this study. However, others have concluded time on CPB is not an influencing factor.

Tissue edema following cardiopulmonary bypass also is a difficult problem and certainly is exacerbated and more difficult to treat in the presence of acute renal failure. The accumulation of water is a consequence believed to be secondary to an inflammatory capillary leak. Total body water itself is increased in these patients, manifested by tissue edema affecting the heart, lungs and brain. Peripheral edema can be severe enough to decrease chest wall compliance, making ventilatory pressure requirements high and chest incision closures difficult. Several factors including young age, long duration and low temperature CPB and low weight have been cited as risk factors. Management of intraoperative techniques, such as ultrafiltration during CPB, has been used as an effective means to reduce postoperative edema. In patients with acute renal failure, dialysis may be started when diuretics are no longer effective and when fluid requirements are excessive.

Treatment
Therapy for acute renal failure is aimed at achieving optimal fluid and electrolyte balance, often in a setting of unstable cardiovascular hemodynamics. This goal becomes more challenging when treatment includes managing nutritional needs of a critically ill patient, and the requirements for drugs that are either excreted by the kidney or possibly nephrotoxic themselves. Interaction between the pediatric nephrologist and the cardiothoracic team, as well as pharmacists and nutritionists, is imperative to achieving a good outcome.

The choice of renal replacement therapy includes the following: peritoneal dialysis, intermittent hemodialysis, continuous hemofiltration or hemofiltration via an extracorporeal membrane oxygenation (ECMO) circuit. All modalities are available at Children's Hospital of Wisconsin, but availability varies at other institutions, based on the center's experience and equipment. Each intervention has advantages and disadvantages. The proper individualized choice should be a joint decision of the team and treating nephrologist.

Peritoneal dialysis is a commonly used modality in the postoperative cardiac patient, especially in patients younger than one year of age. Peritoneal dialysis is performed using a catheter placed in the abdomen for the exchange of custom-made or commercially available dialysate fluids. The peritoneal membrane is a semipermeable membrane lining the abdomen and mesentery, across which osmotic and diffusion gradients can be created and small- and medium-sized molecules - including water - can cross freely. Fluid is allowed to dwell in the abdomen for varying times and can be done continuously either manually by intensive care nurses, or automatically with PD cycler machines. The advantage to this approach is less hemodynamic instability due to the slower, continuous rate of fluid removal over time. In situations of early, low cardiac output, it can be a safe and effective method to control fluid overload. It is not necessarily an inefficient method, however, considering as much as 20-70 ml/kg per day of fluid removal is reported in some series without complication. The main disadvantage of this method is the infectious risk of the catheter. This especially can be a problem in infants in whom catheters leak more easily because they have little subcutaneous fat and muscle to create a snug tunnel for the catheter. The predominant infecting organism is Staphylococcus, but in critically ill patients who already may require broad-spectrum antibiotics, gram-negative organisms or fungal infections can occur. Meticulous exit site care and daily inspection of the peritoneal effluent for early recognition of infection should be standard care in these patients. Another potential disadvantage is that abdominal distension from the dialysis fluid can result in respiratory compromise in patients who already may have significant respiratory distress.

Intermittent hemodialysis is an efficient modality in which the use of high blood flow rates through dialyzer filters with high clearance capacities can adequately control uremia and electrolyte imbalance. Hemodialysis is performed using large, double-lumen catheters to achieve blood flow rates up to approximately four times the patient's weight (ml per minute). These high blood rates are difficult to achieve without good vascular access and pose technical problems in very small infants. Typically, dialysis is performed over shorter periods of two to four hours. The rapid removal of fluid along with the extracorporeal blood in the circuit can result in hemodynamic instability. This often leads to high systemic vascular resistance, tachycardia, or the need for exogenous pressor support to maintain perfusion.

Continuous hemofiltration is a newer modality in the treatment of the pediatric patient. It provides the slow, continuous removal of fluid and solute, performed efficiently like hemodialysis, but ultrafiltration rates can be set lower to that of peritoneal dialysis and be performed over a longer period time. It is indicated in unstable patients with low cardiac output who might not tolerate the fast blood flow rate and ultrafiltration rate of intermittent hemodialysis, or in patients for whom peritoneal dialysis is contraindicated. Patients can remain on this modality for hours, days or, in some cases, weeks if necessary. The development of pump-assisted devices has led to the use of venous access for both inflow and return of blood. This eliminates the need for arterial vascular access required in a "pumpless" system to generate enough filtration pressure through the circuit. Clotting of the filter remains a problem, however. Though anticoagulants can be placed directly into the dialysis circuit, the patient likely will experience systemic anticoagulation. Compared to peritoneal dialysis, continuous hemofiltration requires more technical skill and supervision. Good communication between dialysis staff and intensive care staff needs to be maintained in order for this to be safe and effective.

Finally, hemofiltration through the ECMO circuit can be used and can provide adequate ultrafiltration and solute removal in that population. Dialysate can be added to improve solute clearance, using intravenous fluid pumps at the inflow and outflow dialysate lines to regulate ultrafiltration.

Outcomes
Overall mortality for acute renal failure in pediatric patients is very high at 42.5 percent. In one analysis the rate was much higher in cardiac surgery patients younger than 16 years old (90.9 percent). Rates vary by institution, probably reflecting the differences in decision-making regarding the indications for renal replacement therapy. More consistent is the finding that overall mortality in acute renal failure also is highly dependent on the presence of other complications. Respiratory and neurologic complications appear to increase mortality. In adult studies, pre-surgical nutritional status also influenced outcome. Randomized, prospective studies are needed in the pediatric population to not only fully understand the factors leading to postoperative acute renal failure, but to help differentiate among the many options in renal replacement therapy.

Benzodiazepine Use in the Pediatric Intensive Care Unit

Richard J. Berens, MD, pediatric critical care specialist and chief of staff, Children's Hospital of Wisconsin; assistant professor, Anesthesiology, Medical College of Wisconsin;
Michael T. Meyer, MD, pediatric critical care fellow, Children's Hospital of Wisconsin; instructor, Pediatrics (Critical Care) Medical College of Wisconsin

Note: Because psychotropic medication is a complex topic, opiate use was covered in the October 1999 issue of Heart Matters.

Providing sedation and analgesia in the intensive care setting is driven by a complex mesh of humanitarian goals and physiologic data to improve outcomes and modify stress. Reducing stress remains the central component of modern day anesthesia and sedation in the Pediatric Intensive Care Unit at Children's Hospital of Wisconsin. Primarily, the agents we use are narcotics, benzodiazepines and other sedatives.

Benzodiazepines comprise a family of medications beneficial in the care of children in the PICU. They provide effective antegrade amnesia, anxiolysis, anticonvulsant activity, sedation, hypnosis and skeletal muscle relaxation and share common chemical structure, mechanism of action and side effect profiles. They have no intrinsic analgesic properties and differ in their half-life and metabolic by-products.

The benzodiazepine site of action is the GABA receptor, specifically the a-receptor subunit; GABA is the major inhibitory neurotransmitter within the brain. The receptor functions as an ion channel regulator, permitting the passage of a chloride ion that alters transmembrane potential and "stabilizes" the neuron. This promotes resistance to neuronal excitation which produces sedation, anxiolysis, muscle relaxation and the anticonvulsant activity.

Drug interactions and side effects are similar for all drugs in this classification. One may see potentiation of cardiorespiratory effects from opiods and other CNS depressange agents. Liver blood flow and metabolism may affect the duration of activity. Other drugs dependent on the liver for metabolism may be prolonged. The side effects of benzodiazepines potentially are life threatening and include respiratory depression, hypotension, coma and paradoxical response with hyperexcitation.

Benzodiazepines reduce cerebral metabolism and blood blow in the central nervous system. They alter consciousness in a dose-dependent continuum ranging from anxiolysis to sedation to sleep to general anesthesia. They markedly impair the acquisition of new information ("antegrade amnesia") while maintaining stored information. Benzodiazepines also cause a dose-dependent depression in the respiratory drive. In high doses they can blunt both the hypoxic and the hypercarbic respiratory responses due to hypoventilation. After CV surgery, benzodiazepines produce minimal effects in the euvolemic patient. They reduce both preload and afterload, but cardiac output and arterial blood pressure are minimally affected. However, if the patient is hypovolemic or has catecholamine depletion, significant hypotension may occur, especially when combined with an opiate.

The pharmacokinetic profile of the specific benzodiazepine dictates the use of each particular agent. Pharmocodynamic effects correlate poorly with serum drug levels and may be reflective of the activity of the active metabolites instead. Each drug has varying age-dependent duration of activity from neonates to adults. The specific benefits of each drug are listed below.

Midazolam (Versed) - Four times more potent than diazepam (Valium), it can be painlessly administered intravenously without causing a thrombophlebitis because it is watersoluble. It is characterized by rapid onset and short duration of action because, at physiologic pH, it becomes extremely lipophilic and rapidly crosses the blood-brain barrier. It provides for rapidly titratable sedation and can be administered orally, intranasally and intravenously with reliable absorption and effects.

Diazepam (Valium) - At one time, diazepam was the most commonly used benzodiazepine. It is not very watersoluble, and its solvent vehicle for IV administration includes several organic solvents including propylene glycol. Its poor water solubility makes IM dosing erratic and unpredictable, therefore IV or PO routes are preferred. The IV route also is problematic because it is painful when administered and can cause a thrombophlebitis. It is slowly metabolized in the liver to two active metabolites that are even more slowly cleared, accounting for the long duration of action.

Lorazepam (Ativan) - Is intermediate in action when compared to midazolam and diazepam in onset and duration of action. There are no active metabolites, therefore they will not accumulate in the bloodstream. Lorazepam is metabolized by glucuronidation, a metabolic process well preserved even with hepatic insufficiency.

Feeding the Neonate with Congenital Heart Disease

Sandy Wesolowski, RN, Pediatric Intensive Care Unit, Children's Hospital of Wisconsin

Infants with significant cardiac lesions frequently are difficult to feed. Therefore, about 50 percent are at risk for failure to thrive during their first year of life, even after surgical correction or palliation of their defect.

Many factors contribute to the poor feeding patterns of these infants, including fatigue, working hard to breathe, neurologic insults and delayed gastric emptying.

Barbara Medoff-Cooper, RN, PhD, at the University of Pennsylvania School of Nursing studied these infants using a bottle with a specially designed nipple for measuring the force generated by infants' sucking force. This device has helped identify specific feeding patterns needed for optimal growth. For example, a normal, full-term infant is able to generate a 300 mm to 400 mm Hg suck for a sustained period, while some infants with an interrupted arch are unable to generate 50 mm Hg even in isolated bursts.

It is important to begin enteral feeding as soon as possible after surgery. This not only maximizes the infant's growth potential, but also has been shown to decrease nosocomial infections.

Infants' weights should be monitored daily and their caloric expenditures conserved. For example, using a radiant warmer rather than a crib will decrease the amount of calories the infant needs to expend to maintain body temperature, thereby increasing the calories available for growth.

The early introduction of a pacifier may help stimulate the sucking reflex and make sucking a pleasurable, satisfying experience for the infant. Also, limiting the time food is given by mouth to the first five to 10 minutes of feeding, then gavage feeding the remaining amount, will decrease the infant's workload.

Delayed gastric emptying also potentiates poor feeding. Optimizing gastric emptying with positioning, as well as with drugs such as cisapride, may be necessary. Reducing the amount of narcotics also will decrease factors limiting gastric motility, but care also must be taken to keep the infant comfortable during the postop period. Tylenol can be beneficial.

We can play a vital role in meeting the nutritional needs of these infants. By closely monitoring their caloric needs and minimizing unnecessary energy expenditures we can promote positive growth and development.

Summary of 22q11 deletion

Alexis Sullivan, RN, Pediatric Intensive Care Unit, Children's Hospital of Wisconsin

Dr. Elizabeth Goldmuntz, Cardiology Department at Children's Hospital of Pittsburgh (CHOP), discussed the specific genetic basis of congenital heart disease at the Cardiology 2000 Conference in Orlando in February.

She described the deletion as a loss of the long arm of chromosome 22. Syndromes associated with this deletion include DiGeorge (with conotruncal cardiac defects, for example interrupted aortic arch, truncus arteriosus and tetrology of Fallot) and velocardiofacial syndrome, which can include tetrology of Fallot, ventricular septal defect and right aortic arch.

Goldmuntz studied the frequency of the 22q11 deletion in patients with a conotrucal defect and found that 50 percent (12 of 24) of IAA-type B patients had this deletion. No patients with TGA had the deletion, and only one patient out of 20 tested with DORV had this deletion. Her study and another clinical series reported that arch anomalies were associated with this deletion and the association grew stronger as the severity of the arch defect increased.

Also discussed were noncardiac features reported with 22q11 deletion. In a series of 105 patients who also had palatal findings, including velopharyngeal incompetence, overt or submucosal cleft palate, and immunodeficency - 77 percent had immunodeficiency (67 percent had impaired T-cell production, 23 percent had humoral defects, 18 percent had impaired T-cell function, and 13 percent had IgA deficiency.)

Other noncardiac features associated with 22q11 deletion were hypocalcemia, feeding disorders, growth retardation, speech and learning disabilities, renal and skelatal anomalies and psychiatric/behavioral problems.

According to Goldmuntz, screening for cardiac patients for 22q11 deletion with 5cc of serum (before bypass) using FISH test routinely is recommended at CHOP for patients with interrupted aortic arch or tricuspid atresia. Infants with tetrology of Fallot/pulmonary atresia or TOF and arch/vessel anomalies, as well as patients with cardiac defects who have other 22q11 syndrome features also are screened. If patients are found to have 22q11 deletion, their evaluation includes a genetics consultation, screening of both parents (if both carry the deletion, they have a 50 percent chance of having a second child with the deletion), serum calcium, renal ultrasound (50 percent have renal anomalies), palate and speech evaluation, and psychological and developmental testing.