Feature Article
|
A Complete Primer on Enteral Feeding Improvements in delivery systems for enteral feeding, formulas, and understanding of complications has made the technology for enteral feeding easy to apply. Adequate nutrients can be delivered, and individual tolerance for feeding is acceptable. This review outlines practical tips for prescribing and managing enteral feeding. Long-term enteral feeding is associated with a high complication rate and high mortality, and may not improve underlying clinical conditions.(Annals of Long-Term Care: Clinical Care and Aging 2001;9[1]:41-48)
The indications for enteral nutrition by feeding tube generally fall into one of three categories. First, neuromuscular disease may impair swallowing or gag reflexes. Second, patients’ nutritional needs may not be met through eating alone, as in those with hypermetabolic cancer or cachexia. Finally, an underlying condition may prevent eating, such as in ventilator-dependent or postoperative patients. The only contraindication to enteral feeding is mechanical obstruction of the gut. Diarrhea, vomiting, fistulae, dysmotility, and aspiration represent problems that complicate feeding, but are not contraindications. Patient preferences factor into the decision to use enteral feeding.
Access to the gut can be achieved by nasogastric, nasointestinal, percutaneous gastric, or percutaneous jejunal routes. Each route is associated with benefits and risks, but the percutaneous gastrostomy (PEG) is the most common and arguably the preferred route when the enteral feeding is expected to last for more than 30 days. Most oral medications cannot be used with jejunal intubation (PEJ), and long-term nasogastric intubation (NG) may be uncomfortable.
Enteral feeding tubes come in a variety of specifications. Choices include the external diameter, polyurethane or silicone, weighted or nonweighted tubes, and a variety of different lengths. External diameters range from 8-18 French. Unfortunately, the internal diameter is usually not given. The smaller the internal diameter, the slower the flow, and the longer the feeding cycle necessary to deliver adequate nutrition. Generally, the largest diameter possible should be used for nasogastric enteral tubes, consistent with patient comfort in passing the tube. Small bore tubes are more comfortable but clog more frequently. Larger tubes facilitate administration of medications more effectively. If the patient requires administration of multiple medications through the tube, a larger tube or a PEG tube should be chosen. PEG tubes are large and rarely restrict flow characteristics. PEJ tubes are small and require special handling to prevent clogging; it is very difficult to give medications through a PEJ tube.
Enteral tubes differ in the type of plastic construction. Silicone tubes have thicker walls while polyurethane tubes are thinner. Silicone may cause slower flow and more frequent clogging. Tubes may be weighted or nonweighted. Weighted tubes were formerly thought to be necessary for intestinal intubation. However, there is no research documenting this supposition. Weighted tubes cost more and probably don’t have any advantage over nonweighted tubes. In one study, at four hours, 84% of unweighted tubes had passed into the duodenum, compared with 36% of weighted tubes. At 24 hours, 86% of unweighted tubes had passed into the duodenum, compared to 48% of weighted tubes.1
The lengths of nasogastric tubes vary from 36-45 cm. A tube must be longer than 36 cm to reach the intestine. Longer tubes that cross the pylorus into the duodenum or jejunum have little advantage for reducing complications or improving nutrition. These tubes obviate the advantage of the stomach as a reservoir, require an infusion pump, and are frequently dislodged. One series reported that only 12% of small intestinal tubes were still in use at six weeks, and 20% never reached the duodenum.2
Enteral Formulas
Selecting an enteral product should not be difficult. All formulas contain protein, fat, and carbohydrates. The large number of formulas currently marketed is due to each manufacturer offering a complete line of products for the whole spectrum of nutritional indications. In similar categories, there is very little difference between products. In surveys among nutritionists, cost is cited as the most important factor used by formulary committees in choosing a product.
Products differ in source of protein and in the degree of digestion required (Table I). The choice of a product depends on the condition of the gut and the expected degree of digestion required. As the degree of digestion required lessens, the cost rises considerably. Intact nutrients (blenderized) and milk-based products are the cheapest, and will adequately meet nutritional needs. The most commonly prescribed products, and the largest category, are lactose-free products aimed at preventing lactose-associated diarrhea. Hydrolyzed proteins and crystalline amino acids are used when normal digestive enzymes are not present due to disease or when the absorptive surface of the intestine is altered, in conditions such as pancreatitis or malabsorption. Incomplete amino acids are reserved for patients whose protein must be restricted, in such conditions as renal or hepatic failure.
Simple nutritional calculations are necessary to determine the total volume required to deliver adequate protein, calories, and water (Table II). Proteins are the most critical component. Controversy exists regarding the percentage of the total caloric requirement derived from protein. Generally, formulas contain about 7-16% of total calories from protein. The recommended daily allowance for adults is 0.8 gm/kg/day for protein, which may be too low for older persons. For most institutionalized patients with underlying disease, protein intake should be 1.2 to 1.5 gm/kg/day. However, half of chronically ill elderly persons are unable to maintain nitrogen balance at this level.3 Although adjusting the percentage upward may benefit some special patients, such as those with chronic wounds,4 increasing the total protein percentage may simply supply calories from protein rather than from carbohydrate sources and may dehydrate the patient.5
Caloric requirements range from 25 kcal/kg/day for sedentary adults to 40 kcal/kg/day for stressed adults. Stress generally refers to persons with burns, wounds, cancer, infections, and other similar conditions. In general, caloric requirements can be met at 30-35 kcal/kg/day for elderly patients under moderate stress. Various formulas, including the Harris-Benedict equation, can be used to predict caloric requirements; however, controversy exists over its accuracy in obese or severely malnourished individuals.6 Other formulas have been adjusted for severely stressed hospitalized subjects.7 There is also considerable debate over whether to use ideal body weight or an adjusted body weight in the calculations.
The carbohydrates found in feeding formulas derive from many sources including starch, polysaccharides, di-saccharides, and monosaccharides. The chief difference among the formulas is whether the source includes milk, and thus lactose. The majority of formulas are lactose-free. All products require an intact intestinal brush border for absorption. Most feeding formulas contain approximately 1 kcal/mL. Alternative calories per unit volume are available, ranging from 0.5 to 2.0 kcal/mL. As caloric density increases, gastric motility and emptying decrease, which may increase the risk of aspiration.
Fats are added to formulas for additional calories, for flavor, and for absorption of fat-soluble vitamins. The amount and source of fat differ from product to product. The percentage of fat should be about 30% of calories, but an exact requirement is not known.
Free water requirement is 30-35 mL/kg/day. Each product varies in the amount of free water per unit volume, but in most products it is about 75% of the volume. Free water in the product should be subtracted from the total calculated daily water requirement. Water flushes of the tube can be adjusted to meet free water requirements. Using a low-calorie formula results in a higher volume requirement to meet caloric needs and in increased fluid intake. When 2.0 kcal/cc- density formulas are used, the volume necessary to meet caloric needs decreases, and the amount of free water decreases. This can be useful when fluid restriction is necessary in certain clinical conditions, such as the syndrome of inappropriate secretion of antidiuretic hormone or congestive heart failure requiring fluid restriction.
Products also differ in osmolarity or ionic concentration. Isotonic solutions range from 280-300 mOsm/g. Hypertonic solutions range from 400-1100 mOsm/g. Isotonic formulas can be given at full strength with a rate determined by tolerance. Hypertonic formulas should be initiated at full strength using slow rates (25 mL/hr) or at half strength using 50 mL/hr. It is not clear whether osmolarity makes a significant difference clinically.
Intermittent vs Continuous Feedings
Feedings can be given either intermittently or continuously. The stomach is normally a reservoir for receiving intermittent feedings. For this reason, an intermittent schedule is preferred. There are several advantages to intermittent feedings: they are more convenient for the nursing staff; do not require infusion pumps; permit more patient mobility; better simulate normal feeding patterns with fasting periods; and may be more physiologic.8
If the patient has not been fed in the last five days, feedings should begin as low volume, continuous flow in the range of 25 cc/hr, depending on tolerance. When caloric needs are being met, the schedule can be interrupted for intermittent feedings. The average volume of intermittent feedings should be between 240 and 400 mL. The higher the feeding rate, the higher the residual volume. The gastric residual should be checked prior to each feeding. The caregiver should be concerned if the residual volume is greater than 200 mL for a nasogastric tube, or greater than 100 mL for a gastrostomy tube; however, this does not always necessitate holding feedings. Residual volumes are not reliable if the tube is less than 10 French. Feedings may be given by gravity drainage with a roller clamp, or by a constant infusion pump. There is little data to suggest that infusion pumps are superior;9,10 they are more expensive, and intermittent feeding may be superior for reasons stated above.
Complications
Complications of enteral feeding can be divided into mechanical, gastrointestinal, and metabolic. In the largest study, the overall complication rate was 11.7%.11 The rates in this study were low: gastrointestinal 6.2%, mechanical 3.5%, and metabolic 2.0%. Diarrhea was reported in only 2.3%. Mechanical clogging has been observed only with premixed formulas, with Pulmocare, Ensure, Ensure Plus, Osmolite, and Enrich having the highest potential.12 Vivonex TEN, Hepatic-Aid, Vital, and Citrotein have the lowest potential. The size of the tube is important as well as the viscosity of the formula. The tube should be flushed with at least 30 mL of water every four hours during continuous feeding, after every medication, after intermittent feeding, and after every check for residual. Pills should be dissolved in a solvent before being put through the tube. A pharmacist or drug reference guide should specify the proper solvent. If the drug cannot be dissolved, it may not be possible to use that formulation. Acidity of the medication is related to higher coagulation. A pump should be used with slow or viscous feedings. Water or carbonated cola can be used to clear the tube. Meat tenderizer should not be used.
Pulmonary complications can occur due to improper placement of the tube. Of 100 placements of a nasogastric tube, 19% were neither in the stomach nor the duodenum.13 In elderly patients or those without intact gag reflexes, pneumothorax, pleural penetration, empyema, and bronchopleural fistula may occur more frequently. Using insufflation over the stomach for confirming proper tube placement is frequently misleading. Radiological confirmation of placement should be obtained.
Aspiration of gastric contents is the most serious among the pulmonary complications of tube feedings. As many as 40% of deaths associated with tube feedings result directly from aspiration pneumonia.14 Reported risk factors for aspiration include conditions such as diabetes, pancreatitis, vagotomy, and malnutrition. Formula-associated risks include high-nutrient-density formulas,15 hypo- and hyperosmolar solutions,16 and cold formulas.
There have been suggestions that the type of tube affects the rate of aspiration. PEG tubes were felt to provide more protection than NG tubes, and jejunostomies hypothesized to eliminate the risk. However, aspiration occurs in 44% of NG tube-fed patients and 56% of PEG tube-fed patients in a long-term care setting. Duodenally-placed tubes are not better than PEG tubes.17 In one study, the only risk factor associated with subsequent aspiration in gastrostomy tubes was previous pneumonia.18 Age, mental status, or method of feeding (intermittent vs continuous) was not associated with subsequent risk.
Jejunal tubes placed distal to Treitz’s ligament are theoretically superior to other tubes in the prevention of aspiration and are often recommended as the solution to chronic aspiration. However, there is very little data with which to compare rates of aspiration in jejunal feedings. A literature review found only nine patients with jejunal feedings followed for three months or longer.19 Comparison of PEG and PEJ tubes in another study showed aspiration within 30 days in 5% of PEG patients and 17% of PEJ patients.20 The high rate of PEJ aspiration may occur because these tubes are used in the chronic aspirator. Nevertheless, this strategy is not effective in preventing this serious complication.
Diarrhea is the most frequently reported gastrointestinal complication of tube feeding in most studies. Incidence ranges from 2.3% to as high as 68% in patients in intensive care.21 The pathogenesis of diarrhea is poorly understood. Major studies often report conflicting data in identifying the causes of diarrhea. The factors associated with increased diarrhea include osmolarity, rate of delivery, H2 blockers, antibiotics, and fiber content. Other authors conclude that osmolarity, antibiotics, and H2 blockers are not associated with incidence of diarrhea.22 The discrepancy is likely due to the study design, population, and the definition of diarrhea. Antibiotics have been cited as the most common medications associated with diarrhea, yet the use of multiple antibiotics may be a marker for poor patient condition rather than a cause. In several studies, there has been no association between diarrhea and antibiotic usage. The association with H2 blockers and diarrhea may be related to bacterial overgrowth in patients whose stomach pH exceeds 4 when receiving H2 blockers.23 Drugs containing sorbitol are an often overlooked source of diarrhea.
Malnourished patients have greater fecal energy loss and demonstrated, impaired D-xylose and fat absorption independent of any gastrointestinal tract abnormality. Villous atrophy occurs and results in impaired glucose absorption. Recent work suggests that severely malnourished patients can absorb formulas containing 1 cal/mL.24
Formula osmolarity and rate of feeding have been identified with rate of diarrhea. However, several studies have shown no effect when comparing the osmolarity between 145 and 430 mOsm,25 and no effect on formula composition.26 Formulas high in fiber may decrease the cecal bacterial load or translocation of mesenterial bacteria in rats.27 Whether this is important in humans is not known. Fiber content of formulas probably has little effect on diarrhea related to enteral feeding.
Metabolic complications of enteral feeding include hyperglycemia, hypercapnia, and electrolyte abnormalities. Complications are seen more often with high-caloric-density formulas or in patients with diabetes. Special formulas are available for diabetic patients, but there is limited data on their effectiveness. Administering hyperosmolar formulas to diabetic patients may lead to hyperosmolar nonketotic coma. High-carbohydrate concentrations may increase respiratory quotients and increase carbon dioxide production. Tube-feeding syndrome is characterized by dehydration, hypernatremia, hyperchloremia, and azotemia. The most common cause of this syndrome is high-protein formulas with low water intake. Other syndromes occur from an imbalance of water or salt, from diarrhea, or from renal dysfunction.
Ethical Considerations
Whether to begin enteral feeding is often an ethical dilemma. Whether or not enteral feeding improves outcome is problematic. It remains axiomatic that clinical condition is not improved by malnutrition; without adequate intake of nutrients, death will occur sooner. It is less clear, however, whether the outcome of the underlying clinical condition is improved by long-term enteral nutrition. The complication rate for enteral feeding is high. In 100 consecutive patients with PEG placement, 41% died within 30 days, 4% of whom died directly from the procedure.28 The 30-day mortality rate in subjects admitted from a nursing home for PEG placement (87% for dementia) was 39.5%.29
Enteral feeding with a PEG tube can be well tolerated for long periods of time. In 210 subjects followed for four years, acceptability reported by patients was excellent in 83%, sufficient in 15%, and poor in only 2%. However, the one-year survival rate in this study was only 34%.30
In general, much more data exists for short-term enteral feedings than for long-term enteral feedings. A review of the literature will allow several conclusions. When enteral and parenteral nutrition have been prospectively compared in patients who could receive either, there was no apparent advantage of one route over the other. Morbidity was higher with parenteral feedings. No consensus exists as to whether perioperative nutritional support reduces mortality or complications to the level of well-nourished patients.31
Very little data is available regarding the outcome of enteral feedings in long-term care settings. Age alone is not a factor in clinical outcome. Kaminski et al32 studied 102 consecutive patients who were nutritionally supported to determine the effect of age on response and outcome of hospital stay. Subjects were divided into two groups: group 1 (N=37) was under the age of 65 years and group 2 (N=65) was over the age of 65 years. Group 1 was fed 37.8±15 days, and group 2 was fed 44.8±28 days. Crude mortality was 13% in group 1 and 35% in group 2. Improvement or maintenance of somatic parameters occurred in 89% of group 1 and in 83% of group 2 (p > 0.1). Improvement or maintenance of visceral parameters occurred in 78% of group 1 and in 72% of group 2 (p > 0.1). Although improvement was similar, the mortality was higher in the elderly patients.
The effects of enteral feeding were prospectively studied for 11 months in 70 tube-fed subjects in a long-term care facility.10 Indications for enteral feeding included refusal to swallow (50%), dysphagia without obstruction (47%), and esophageal obstruction (3%). Problems that occurred within the first two weeks in the NG tube-fed patients included agitation and self-extubation (N=38, 67%), and aspiration pneumonia (N=23, 43%). Early problems that occurred in the gastrostomy tube-fed patients included aspiration pneumonia (N=9, 56%), tube dysfunction (N=8, 50%), and agitation and self-extubation (N=7, 44%). Later problems included aspiration pneumonia in the NG tube-fed group (N=24, 44%), versus 56% of the gastrostomy tube-fed group (N=9), and tube dysfunction in 38% (N=6) of the gastrostomy tube-fed group. Late extubation was limited to the NG tube-fed group (N=21, 38%). Responses to tube feedings were observed in 56 patients for 11 months. Weight was stable for six months, but weight loss was more apparent beyond six months. Only 6% of the patients gained weight at any time. Albumin was stable for at least one month after starting enteral feedings. Wide fluctuations occurred, but the mean albumin remained stable during the study. The mean end-of-study albumin was not normalized, however. Hemoglobin also remained stable during the 11-month follow-up, but did not increase over the mean at the start of the study. Death occurred in 28 patients, usually between the second and sixth months, with a mortality rate of 46%. Based on this data, continuation of tube feedings for long periods of time is associated with a high frequency of complications. Aspiration probably contributed to death in 40% of patients. Body weight may be maintained for six months, followed by continued loss of weight.
Two reports using an administrative database during six months of follow-up reported that tube feeding was not associated with healing of preexisting pressure sores,33 or with protection from new pressure sores.34 In an observational study of enterally-fed patients, the prevalence of pressure ulcers was not affected.35
Survival does not appear to be affected by enteral feeding. Of 1386 nursing home residents older than 65 years with recent progression to severe cognitive impairment, 9.7% had a feeding tube placed. Survival for a period of 24 months was the same for residents who were and were not tube-fed.36
The ethical dilemma that results from this data is obvious. Intuitively, no clinical condition is improved by malnutrition, and without adequate nutritional intake death will occur. However, clinical data suggest that the complication rate of enteral feeding is high and that clinical outcome is often not improved by enteral feeding.
Conclusions
Improvements in delivery systems for enteral feeding, formulas, and understanding of complications has made the technology for enteral feeding easy to apply. Adequate nutrients can be delivered, and individual tolerance for feeding is acceptable. The remaining question is when to apply the technology. Formula selection should be as simple as possible. Aspiration and other early complications are a serious risk and are not diminished by route of feeding. Long-term enteral feeding is associated with a high complication rate, high mortality, and low effectiveness.
--------------------------------------------------------------------------------
References
1. Lord LM, Wiser-Maimone A, Pulhamus M, et al. Comparison of weighted vs unweighted enteral feeding tubes for efficacy of transpyloric intubation. J Parenter Enteral Nutr 1993;17:271-273.
2. Lipman TO. The fate of enteral feeding tubes. Nutrition Supplement Service 1983;3:71.
3. Gersovitz M, Motil K, Munro HN, et al. Human protein requirements: Assessment of the adequacy of the current recommended dietary allowance for dietary protein in elderly men and women. Am J Clin Nutr 1982;35:6-14.
4. Breslow RA, Hallfrisch J, Guy DG, et al. The importance of dietary protein in healing pressure ulcers. J Amer Geriatr Soc 1993;41:357-362.
5. Klein CJ, Stanek GS, Wiles CE III. Overfeeding macronutritents to critically ill adults: Metabolic complications. J Am Diet Assoc 1998;98:795-806.
6. Choban PS, Burge JC, Flanobaum L. Nutrition support of obese hospitalized patients. Nutrition in Clinical Practice 1997;12:149-154.
7. Ireton-Jones CS. Evaluation of energy expenditures in obese patients. Nutrition in Clinical Practice 1989;4:127-129.
8. Rombeau JL. Nasoenteric tube feeding. In: Rombeau JL, Caldwell MD, eds. Clinical Nutrition. vol 1. Philadelphia, PA: WB Saunders; 1984: 261.
9. Kocan MJ, Hickisch SM. A comparison of continuous and intermittent enteral nutrition in NICU patients. J Neurosci Nurs 1986;18:333-337.
10. Ciocon JO, Galindo-Ciocon DJ, Tiessen C, Galindo D. Continuous compared with intermittent tube feeding in the elderly. J Parenter Enteral Nutr 1992;16:525-528.
11. Cataldi-Betcher EL, Seltzer MH, Slocum BA, Jones KW. Complications occurring during enteral nutrition support: A prospective study. J Parenter Enteral Nutr 1983;7:546-552.
12. Mancuard RP, Perkins AM. Clogging of feeding tubes. J Parenter Enteral Nutr 1988; 12:403-405.
13. Benya R, Langer S, Mobarhan S. Flexible nasogastric feeding tube tip malposition immediately after placement. J Parenter Enteral Nutr 1990;14:108.
14. Ciocon JO, Silverstone FA, Graver LM, Foley CJ. Tube feedings in elderly patients: Indications, benefits, and complications. Arch Intern Med 1988;148:429.
15. Hunt JN, Stubbs DF. The volume and energy content of meals as determinations of gastric emptying. J Physiol 1974;245:209.
16. Davenport HW. Physiology of the Digestive Tract. 4th ed. Chicago, IL: Year Book Medical Publishers; 1977.
17. Gustke R, Varme R, Soergel K. Gastric reflux during perfusion of the proximal small bowel. Gastroenterology 1970;59:890-895.
18. Cogen R, Weinryb J. Aspiration pneumonia in nursing home patients fed via gastrostomy tubes. Am J Gastroenterol 1989;84:1509-1512.
19. Lazarus BA, Murphy JB, Culpepper L. Aspiration associated with long-term gastric versus jejunal feeding: A critical analysis of the literature. Arch Phys Med Rehabil 1990;71:46-53.
20. Wolfsen HC, Kozarek RA, Ball TJ, et al. Tube dysfunction following percutaneous endoscopic gastrostomy and jejunostomy. Gastrointest Endosc 1990;36:261-263.
21. Kelly TWJ, Patrick MR, Hillman KM. Study of diarrhea in critically ill patients. Crit Care Med 1983;11:7-9.
22. Benya R, Mobarhan S. Enteral alimentation: Administration and complications. J Am Coll Nutr 1991;10:209-219.
23. Hillman KM, Riordan R, O’Farrell SM, Tabaqchali S. Colonization of the gastric contents in critically ill patients. Crit Care Med 1982;10:444-447.
24. Benya R, Montegeudo J, Zarling E, Mobarhan S. Carbohydrate and protein absorptive efficacy in well nourished and malnourished patients. J Am Coll Nutr 1991;10:51-56.
25. Keohane PP, Attrill H, Love M, et al. Relationship between osmolality of diet and gastrointestinal side effects in enteral nutrition. BMJ 1984;288:678-680.
26. Zarling EJ, Parmar JR, Mobarhan S, Clapper M. Effect of enteral formula rate, osmolality, and chemical composition upon clinical tolerance and carbohydrate absorption in normal subjects. J Parenter Enteral Nutr 1986;10:588-590.
27. Alverdy JC, Aoys E, Moss GS. Effect of commercially available chemically defined liquid diets on the intestinal microflora and bacterial translocation from the gut. J Parenter Enteral Nutr 1990;14:1-6.
28. Oyogoa S, Schein M, Gardezi S, Wise L. Surgical feeding gastrostomy: Are we overdoing it? Journal of Gastrointestinal Surgery 1999;3(2):152-155.
29. Abuksis G, Mor M, Segal N, et al. Percutaneous endoscopic gastrostomy: High mortality rates in hospitalized patients. Am J Gastroenterol 2000;95:128-32.
30. Loser C, Wolters S, Folsch UR. Enteral long-term nutrition via percutaneous endoscopic gastrostomy (PEG) in 210 patients: A four-year prospective study. Dig Dis Sci 1998;43:2549-2557.
31. Campos ACL, Meguid MM. A critical appraisal of the usefulness of perioperative nutritional support. Am J Clin Nutr 1992;55:117-130.
32. Kaminski MV, Nasr NJ, Freed BA, Sriram K. The efficacy of nutritional support in the elderly. J Am Coll Nutr 1982;1:35.
33. Berlowitz DR, Brandeis GH, Anderson J, Brand HK. Predictors of pressure ulcer healing among long-term care residents. J Am Geriatr Soc 1997;45:30-34.
34. Berlowitz DR, Ash AS, Brandeis GH, et al. Rating long-term care facilities on pressure ulcer development: Importance of case-mix adjustment. Ann Intern Med 1996;124:557- 563.
35. Henderson CT, Trumbore LS, Mobarhan S, et al. Prolonged tube feeding in long-term care: Nutritional status and clinical outcomes. J Am Coll Nutr 1992;11:309-325.
36. Mitchell SL, Kiely DK, Lipsitz LA. The risk factors and impact on survival of feeding tube placement in nursing home residents with severe cognitive impairment. Arch Intern Med 1997;157:327-332.
--------------------------------------------------------------------------------
Dr. Thomas is Professor, Internal Medicine, Division of Geriatric Medicine, St. Louis Health Sciences Center, St. Louis, MO. Address for correspondence: David R. Thomas, MD, Division of Geriatric Medicine, St. Louis Health Sciences Center, St. Louis University, 1402 S Grand Blvd, Room M238, St. Louis, MO 63104. Annals of Long-Term Care - ISSN: 1524-7929 - Volume 9 - Issue 01 - January 2001 |