NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Holzheimer RG, Mannick JA, editors. Surgical Treatment: Evidence-Based and Problem-Oriented. Munich: Zuckschwerdt; 2001.

Cover of Surgical Treatment

Surgical Treatment: Evidence-Based and Problem-Oriented.

Show details

The immunocompromised patient

, M.D.

Author Information

Management of the immunocompromised patients is a vast topic, impossible to cover comprehensively in a brief review. Therefore, salient points related to four common and sometimes confusing entities - adrenal insufficiency, diabetes mellitus, febrile neutropenia, and the human immunodeficiency virusinfected patient - are provided.

Adrenal insufficiency

Considering the broad range of indications for glucocorticoid therapy, it is certain that every general surgeon will encounter a patient who takes corticosteroids (either orally or by inhalation) for underlying diseases as diverse as asthma, connective tissue disorders, malignant disease, or a solid organ transplant. Recognizing that exogenous steroids (even low-dose inhaled drug) can suppress the hypothalamic-pituitary-adrenal (HPA) axis, and consequently the counterregulatory stress hormone response to illness and tissue injury, it has been believed for more than 40 years that stressed patients who take glucocorticoids chronically should receive “stress” steroid hormone replacement, often in large doses. Plasma ACTH and cortisol concentrations increase in response to surgery and trauma, but there is considerable patient-to-patient variability, partly due to the potential suppressive effects of effects of anesthesia, opioids, antihypertensive agents, age, infection, or other variables.

There is remarkably little evidence to support the widespread use of “stress-dose” steroids to steroid-dependent stressed patients. According to an analysis by Kehlet (reported in (1)), three of the 57 initial reports of postoperative death or hypotension putatively ascribed to adrenocortical insufficiency fulfilled conventional diagnostic criteria. Moreover, a review of more than 72,000 operations in three large series indicated that the incidence of perioperative adrenal insufficiency is 0.1% at most (1). The problem does exist, but the incidence is far less than the multitude of patients who receive “stressdose” steroids each year. Prospective studies indicate that the problem is rare even in patients at risk. Bromberg et al. (2) studied 40 renal allograft recipients who were readmitted for surgery or an acute illness. None developed adrenal insufficiency despite receiving no increase in their baseline steroid dose. Glowniak and Loriaux (3) randomized 18 steroid-dependent patients (> 7.5 mg prednisone/day) who were proved to have suppressed HPA axis function by provocative testing to receive either only their baseline steroids or supplemental cortisol in the perioperative period after major surgery. Non-supplemented patients did not develop tachycardia or hypotension caused by inadequate glucocorticoid levels in the perioperative period.

If there is any question, the HPA axis can be tested with an ACTH stimulation test. A baseline cortisol concentration is determined, followed by administration of 250 μg synthetic 1-24ACTH. Interpretation is controversial. The 30–60 min post-stimulation cortisol concentration should increase by ≥ 7 mg/dL, but some investigators believe that the latter determination must be also > 20 mg/dL. Neither method is superior, but some investigators suggest that the 250 μg dose is insensitive to subtle derangements, and can be replaced with a 1 μg dose (4).

Normal daily cortisol production averages 35 mg/day. Estimates of maximal production following major surgery range from 60–310 mg/day in individual patients, with a mean value of 137 mg/day (1). It is seldom that even a maximally-stressed patient would require more than 150 mg/day of exogenous hydrocortisone or equivalent. Salem et al. (1), based on expert opinion (there are no Class I data), recommended that patients on chronic glucocorticoid therapy receive an additional 25 mg hydrocortisone equivalent for minor surgical stress (e.g., inguinal hernia repair). The recommend supplementary dose for moderate surgical stress (e.g., partial colectomy) is 50–75 mg hydrocortisone equivalent for 1–2 days. Maximal surgical stress (e.g., cardiopulmonary bypass) would require no more than 100–150 mg hydrocortisone equivalent for 2–3 days.

The other important aspect of secondary adrenal cortical insufficiency is that patients with previously normal HPA axis function can develop adrenal insufficiency while under severe stress (5). The incidence of adrenal insufficiency was 0.66% in a prospective study of critically ill surgical patients (6), but increased to 11% in patients > age 55 who stayed in the ICU for > 14 days. Diagnostic suspicion must be high, but provocative testing can make the diagnosis easily. Whether intervention in diagnosed cases improves outcome remains unproved.

Diabetes mellitus

Patients with diabetes are at higher risk for a need for surgical intervention than their non-diabetic counterparts. Perioperative management of the patient with diabetes is commonplace and potentially complex, but again, there are few Class I data available to inform management. Surgical stress increases circulating glucose concentration acutely via catecholamine-mediated mobilization of glycogen stores, and also induces transitory insulin resistance. After elective surgery, the effect usually subsides after about 48 hours (7), but sepsis or other prolonged stresses increase the magnitude and duration. Longerterm, catabolism of lean tissue provides substrate for gluconeogenesis in the fasted, stressed patient. Early work suggested that the administration of small amounts of dextrose had a protein-sparing effect, but the effect is inconsequential clinically in stressed patients. The insulin resistance is important because hyperglycemia is common perioperatively in patients who receive full nutritional support, even if they were not insulin-dependent beforehand. Fortunately, the effect is usually transitory. Caution is necessary when managing patients with severe liver disease, because glycogen stores and the gluconeogenic response may be inadequate and hypoglycemia may be precipitated by disease or drug therapy.

Working against this effect is that patients are usually held NPO prior to surgery. For the diabetic patient on oral hypoglycemic agents or insulin, there is also the danger of hypoglycemia in the perioperative period. In type-1 diabetic patients, long-acting insulins, such as ultralente of animal origin should be stopped preoperatively and substituted by protamine and lente (regular) insulins (8). If an evening dose is taken, it should be reduced (or omitted) if the patient will be NPO without intravenous dextrose the night before surgery (typical for ambulatory or sameday surgery patients). In type-2-diabetic patients, long-acting sulfonylurea drugs such as chlorpropamide should be stopped and substituted by short-acting agents. Metformin must always be stopped, ideally no later than 48 hours prior to surgery, because severe metabolic acidosis can precipitated. Type-2-diabetic patients with marked hyperglycemia under oral treatment should be switched to insulin before operation. Absent any clinical risk factor for coronary artery disease other than diabetes, routine preoperative provocative cardiac screening is unnecessary (9).

Tight perioperative control of blood glucose for the diabetic patient is a matter of controversy (10). Diabetes per se is not a risk factor for post-operative morbidity or mortality, after adjustment for co-morbidities such as atherosclerosis. However, controlled studies have shown the advantages of tight control in two acute complications of atherosclerosis, i.e. myocardial infarction and cerebrovascular accident (when focal or global cerebral ischemia is a planned part of the operation) (10). There is also evidence that good glucose control (< 220 mg/dL) can reduce the risk of nosocomial infection (11).

The magnitude of surgery influences the amount of insulin required during surgery (12). The insulin requirements in diabetic patients during surgery vary from 0.25–0.40 U per gram glucose in normal weight patients, 0.4–0.8 U per gram glucose in case of obesity, liver disease, steroid therapy or sepsis, to 0.8–1.2 U per gram glucose in patients undergoing cardiopulmonary bypass surgery. However, there is no agreement as to the optimal method of insulin administration (13), and careful administration of bolus dose of insulin (-10 U q2h) may be comparable to the popular technique of continuous infusion during surgery (~ 1.25 U/hr) (14). In the immediate postoperative period, good control of blood glucose can simplify fluid management by eliminating osmotic diuresis as a confounding variable, and may also reduce fluctuations in mental status. Continuous infusions of insulin are increasingly popular for tight control, but must always be accompanied by an infusion of dextrose to reduce the risk of catastrophic insulin overdosage. High-quality bedside glucose monitors also allow the careful titration of small boluses of intravenous insulin as needed. Subcutaneous insulin can be absorbed erratically owing to the fluid shifts that occur in the immediate postoperative period, therefore the intravenous route is most reliable.

Febrile neutropenia

Clinical use of powerful immunotherapies and multidrug chemotherapeutic regimes have made infections complications in neutropenic patients a major problem. Profound neutropenia may also be caused by adverse drug reactions. Infection is the leading cause of morbidity and mortality in patients with cancer, and neutropenia is a critical cofactor. Febrile neutropenia is now sufficiently common that virtually every busy general surgeon will encounter cases in practice. Both the magnitude and duration of neutropenia are independent risk factors for infection (15). Mortality from bacteremia has been reduced by the use of empiric antibiotics from 90% in the 1950's to 10–30% currently, depending on the causative organism (16). Mortality is also reduced in patients who are neutropenic for less than seven days.

Historically, aerobic gram-negative bacilli - E. coli, Klebsiella, Enterobacter, Acinetobacter and especially Pseudomonas - were the most frequent pathogens, and high resistance rates often mandated double gram-negative antibiotic therapy. Recently, gram-positive pathogens-methicillin-resistant staphylococci (both S. aureus and S. epidermidis), β-hemolytic streptococci, diptheroids, and clostridial organisms - are increasingly common (17) and have supplanted the gram-negative bacteria in may centers. Invasive fungal infections are an so increasing in incidence, especially with species that are likely to be resistant to fluconazole (e.g., Candida glabrata, C. krusei, C. tropicalis) (18). Despite this change in epidemiology, there is a trend in the literature toward empiric antibiotic monotherapy for these patients (19).

Fever in these patients, most often caused by infection, is defined as a single temperature > 38.3 °C absent another obvious cause (e.g., blood transfusion), or a persistent temperature > 38.0 °C for > 4 hours in a 24 h period (19). Neutropenia is generally defined as an absolute neutrophil count (ANC) < 500/mm3. The decision to obtain cultures and start empiric antibiotics is based on a careful history and physical examination. The surgeon must be alert to the possibility that a non-bacteremic infection may be present, especially within the abdomen. Cancer chemotherapy patients may develop acute acalculous cholecystitis, or something as mundane as acute appendicitis. Specific to the neutropenic patients is the entity of neutropenic enterocolitis (20). The spectrum of disease is wide in neutropenic enterocolitis, and the presenting symptoms are non-specific. The diagnosis is best confirmed by computed tomography. Most cases respond well to antibiotics and bowel rest, and do not require surgical intervention.

The choice of empiric antibiotics is guided by several principles (19). The antibiotics must achieve bactericidal concentrations rapidly. Broad-spectrum coverage must account for potential skin/soft tissue, intestinal, and catheter-related sources of infection and the known microbiologic possibilities. Risk stratification schemes are now allowing certain lowrisk patients to be treated with oral antibiotics as outpatients. For inpatients, suitable combination regimes include an anti-pseudomonal penicillin (e.g., piperacill in-tazobactam, ticarcillin-clavulanate) or cephalosporin (e.g., cefepime, ceftazidime) plus an aminoglycoside. Alternatives include an anti-pseudomonal β-lactam drug plus another β-lactam agent, or a combination of anti-pseudomonal β-lactam/aminogylcoside/glycopeptide (e.g., vancomycin). Several clinical studies (19) suggest that monotherapy can be as effective as combination therapy, with lower cost and possibly greater tolerability. Monotherapeutic agents supported by Class I data include ceftazidime, cefepime, imipenem/cilastatin, meropenem, and piperacillin-tazobactam. Current published guidelines consider monotherapy and combination therapy to be equivalent (15).

If the patient responds promptly (within 72 h) and no organism is identified, therapy should continue for at least seven days, or until neutropenia resolves (ANC > 1000/mm3) with no further fever. Persistent fever and no organism should prompt evaluation for a non-infectious cause of fever, an occult resistant organism, or inadequate therapy. The antibiotics need be changed only for evidence of disease progression or chemical deterioration. Addition of vancomycin may be considered after 72 h, and persistent fever for more than seven days is an indication for empiric antifungal therapy with either fluconazole or amphotericin B. Regardless of the clinical scenario, the use of granulocyte colony stimulating factors as adjunctive therapy until the ANC recovers reduces morbidity but not mortality in these patients (21).

Human immunodeficiency virus infection

Effective nucleoside anti-retroviral agents and protease inhibitors have improved dramatically the outlook of HIV-infected patients, and altered completely the nature of the surgical care that these patients require (22). Formerly, opportunistic infections and catastrophic complications such as intestinal perforation secondary to either infection or neoplasia were commonplace (23, 24), but no longer. Now, more routine complications such as acute appendicitis and cholecystitis are common, and outcomes are comparable to non-infected patients. If the patient's CD4+ count is not depressed, the surgical outcome should be good. HIV-infected patients have relatively preserved immunity against conventional bacterial infections, and there is no evidence to suggest that HIV-infected patients with surgical infections caused by bacteria need modified antibiotic regimes or a prolonged duration of therapy.

Acute cholecystitis is now the most common indication for celiotomy in patients with AIDS. Two forms of acalculous biliary disease occur in patients with AIDS. One form is cholestasis similar to sclerosing cholangitis, which can be impossible to distinguish from bacterial cholangitis in an acutely jaundiced patient. Alternatively, AIDS patients present with acute calculous or acalculous cholecystitis. Absent gallstones, both of these syndromes have been associated with Cytomegalovirus (CMV) infection, or infection with Crytosporidium or microsporidial protozoa; these opportunistic infections account for the majority of cases of acalculous cholecystitis in patients with AIDS. In the presence of gallstones, however, the pathogens are the typical bacteria of acute cholecystitis.


Salem M, Tainsh R E Jr, Bromberg J. et al. Perioperative glucocorticoid coverage. A reassessment 42 years after emergence of a problem. Ann Surg. (1994);219:416–425. [PMC free article: PMC1243159] [PubMed: 8161268]
Bromberg J S, Alfrey E T, Barker C F. et al. Adrenal suppression and steroid supplementation in renal transplant recipients. Transplantation. (1991);51:385–390. [PubMed: 1847249]
Glowniak J U, Loriaux D L. A doubleblind study of perioperative steroid requirements in secondary adrenal insufficiency. Surgery. (1997);121:123–129. [PubMed: 9037222]
Merry W H, Caplan R H, Wickus G G. et al. Postoperative acute adrenal failure caused by transient corticotropin deficiency. Surgery. (1994);116:1095–1100. [PubMed: 7985093]
Soni A, Pepper G M, Wyrinski P M. et al. Adrenal insufficiency occurring during septic shock: Incidence, outcome, and relationship to peripheral cytokine levels. Ann J Med. (1995);98:266–271. [PubMed: 7872343]
Barquist E, Kirton O. Adrenal insufficiency in the surgical intensive care unit patient. J Trauma. (1997);42:27–31. [PubMed: 9003254]
Kaufman F R, Devgan S, Roe T F, Costin G. Perioperative management with prolonged intravenous insulin infusion versus subcutaneous insulin in children with type I diabetes mellitus. J Diabetes Complications. (1996);10:6–11. [PubMed: 8639976]
Peters A, Kerner W. Perioperative management of the diabetic patient. Exp Clin Endocrinol Diabetes. (1995);103:213–218. [PubMed: 7584526]
Zarich S W, Cohen M C, Lane S E. et al. Routine perioperative dipyridamole 201T1 imaging in diabetic patients undergoing vascular surgery. Diabetes Care. (1996);19:355–360. [PubMed: 8729159]
Vanhaeverbeek M. Peri-operative care: management of the diabetic patient. A novel controversy about tight glycemic control. Acta Clin Belg. (1997);52:313–319. [PubMed: 9489126]
Pomposelli J J, Baxter J K III, Babineau T J. et al. Early postoperative glucose control predicts nosocomial infection rate in diabetic patients. JPEN J Parenteral Enteral Nutr. (1998);22:77–81. [PubMed: 9527963]
Raucoules-Aime M, Roussel L J, Rossi D. et al. Effect of severity of surgery on metabolic control and insulin requirements in insulin-dependent patients. Br J Anaesth. (1995);74:231–233. [PubMed: 7696075]
Clark J D, Currie J, Hartog M. Management of diabetes in surgery: A survey of current practice by anaesthetists. Diabet Med. (1992);9:271–274. [PubMed: 1576811]
Raucoules-Aime M, Ichai C, Roussel L J. et al. Comparison of two methods of i.v. insulin administration in the diabetic patient during the perioperative period. Br J Anaesth. (1994);72:5–10. [PubMed: 8110550]
Hughes W T, Armstrong D, Bodey G P. From the working committee, Infections Diseases Society of America. Guidelines for the use of anti-microbial agents in neutropenic patients with unexplained fever. J Infect Dis. (1990);161:381–396. [PubMed: 2179420]
Klasterly J. A review of chemoprophylaxis and therapy of bacterial infections in neutropenic patients. Diagn Microbiol Infect Dis. (1989);12 (suppl 4):201S–207S. [PubMed: 2686921]
Pizzo P A. Management of fever in patients with cancer and treatment-induced neutropenia. N Engl J Med. (1993);328:1323–1332. [PubMed: 8469254]
Mennier F, Aoun M, Bitar N. Candidenia in immunocompromised patients. Clin Infect Dis. (1992);14 (suppl 1):S 120–S125. [PubMed: 1562683]
Hathorn J W, Lyke K. Emperical treatment of febrile neutropenia: Evolution of current therapeutic approaches. Clin Infect Dis. (1997);24 (suppl 2):S 255–S265. [PubMed: 9126701]
Song H K, Kreisel D, Canter R. et al. Changing presentation and management of neutropenic enterocolitis. Arch Surg. (1998);133:979–982. [PubMed: 9749851]
Gilbert C, Meisenberg B, Vredenburg J. et al. Sequential prophylactic oral and empiric once-daily parenteral antibiotics for neutropenia and fever after high-dose chemotherapy and autologous bone marrow support. J Clin Oncol. (1994);12:1005–1011. [PubMed: 8164024]
Flum D R, Wallack M K. The surgeon's database for AIDS: A collective review. J Am Coll Surg. (1997);184:403–412. [PubMed: 9100689]
Mueller G P, Williams R A. Surgical infections in AIDS patients. Am J Surg. (1995);169 (suppl 5A):34S–38S. [PubMed: 7755166]
Lowy A M, Barie P S. Laparotomy in patients infected with human immunodeficiency virus: Indications and outcome. Br J Surg. (1994);81:942–945. [PubMed: 7922084]
Copyright © 2001, W. Zuckschwerdt Verlag GmbH.
Bookshelf ID: NBK6938


  • PubReader
  • Print View
  • Cite this Page

Related information

  • PMC
    PubMed Central citations
  • PubMed
    Links to PubMed

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...