effects of concurrent sepsis with clinically significant pulmonary embolic disease

Effects of Concurrent Sepsis with Clinically Significant Pulmonary

Embolic Disease

Britton West, MD, Louisville, Kentucky

Richard N. Garrison, MD, Louisville, Kentucky

Lewis M. Flint, Jr, MD, Louisville, Kentucky

Pulmonary embolism remains an important com-

plication of major operative procedures despite sig-

nificant advances in diagnosis and treatment which have occurred since the original description of

thromboembolism by Virchow [I] and the recogni- tion by Sevitt and Gallagher [2] that pulmonary

embolus often follows deep venous thrombosis. The

diversity of pulmonary complications which may be

encountered in postoperative patients makes accurate diagnosis difficult. Atelectasis, pneumonia, and

adult respiratory distress syndrome can mimic the

typical signs, symptoms, and roentgenographic findings of pulmonary embolism. Identification of

specific patients who are at high risk for the devel-

opment of pulmonary embolism has been difficult.

Furthermore, the benefit of prophylactic minidose heparin in preventing pulmonary embolism has not,

as yet, been confirmed, because the overall incidence of pulmonary embolism is less than 2 per cent [3,4] and because very large numbers of patients must be

studied [5] to achieve valid results. Further confusion arises from the clinician’s inability to accurately

diagnose pulmonary embolism using noninvasive

technics, the variability of heparin dosage schedules, and conflicting reports citing the efficacy of other

anticoagulant drugs such as aspirin [6], crystalline

sodium warfarin [2,5], and dextran [ 71. Statistically significant relationships between

postoperative pulmonary embolism and associated chronic disease are well known. Important factors include age, obesity, atherosclerotic cardiovascular disease, congestive heart failure, neoplasm, prior thromboembolism, and lower extremity venous dis-

From the Department of Surgery, University of Louisville School of Medicine, Health Sciences Center, Louisville, Kentucky.

Reprint requests should be addressed to Lewis M. Flint, Jr, MD, Depart- ment of Surgery, University of Louisville School of Medicine. Health Sciences Center, P.O. Box 35260, Louisville, Kentucky 40232.

ease. Other factors associated with increased risk for

pulmonary embolism are amputation, lower ex-

tremity trauma, pelvic operation, and oral contra-

ceptives (in women). Prophylactic heparin is rec- ommended for such “high risk” groups. Certain pa-

tients, however, such as those undergoing open

prostatectomy [8] or total hip replacement [9] are at high risk for pulmonary embolism but are not pro-

tected by prophylactic heparinization.

Material and Methods

We recently examined the records of 10,000 patients operated on during the period 1971 to 1976, seeking factors associated with increased risk for postoperative pulmonary embolism. Operating records, autopsy records, radiology reports, and proceedings of audit committees were screened to tabulate instances of fatal or significant pulmonary embolism.

Methods of diagnosing pulmonary embolism included autopsy specimens, pulmonary angiographic evaluation, serial perfusion lung scans, and clinical presentation in- cluding electrocardiogram and laboratory data. Strict diagnostic criteria were emphasized so that only clinically significant, potentially fatal, or documented fatal emboli were included. When a lung scan was used as the primary diagnostic technic, clinical symptoms of pulmonary embolism and a negative chest roentgenogram or a chest x-ray film compatible with pulmonary embolism were re- quired.

Strict diagnostic standards were likewise utilized in determining associated disease processes. The criteria for diagnosing infection were: (1) identification of the source of infection; (2) a culture positive for a specific organism; and (3) clinical signs of infection, such as leukocytosis, fever, and purtilent drainage. Heart disease was diagnosed using clinical and electrocardiographic evidence and clinical criteria of the New York Heart Association, while any patient more than 50 pounds over ideal body weight was considered obese.

356 The American Journal of Surgery

TABLE I Factors Associated with Pulmonary Embolus

Factor Incidence of Association

Infection Heart disease Obesity Age >60 yr Trauma to lower extremity Thrombophlebitis Amputation Neoplastic disease Previous thrombophlebitis

39% 36% 32% 29% 17% 14% 10% 5.9% 2.5%

Results

The various diagnostic means disclosed 21 cases of fatal postoperative and 20 cases of significant nonfatal postoperative pulmonary embolism. These patients ranged in age from nineteen to eighty years (mean, 51 years) and consisted of twenty-five women and sixteen men.

The overall incidence of fatal and significant postoperative pulmonary embolism was approximately 0.41 per cent. Factors most commonly associated with pulmonary embolus on our surgical service were heart disease, obesity, infection, and age more than sixty years. (Table I.) Only two patients had cancer (pancreas and colon) at the time of pulmonary embolism, and only one patient had a previous episode of thrombophlebitis. Six of the forty- one patients had clinical deep vein thrombophlebitis at the time of pulmonary embolism. Patients more than sixty years old were most likely to suffer a fatal postoperative pulmonary embolism (70 per cent); such fatalities were possibly related to lower cardiac reserve. Younger patients were more likely to survive the episode (78 per cent). In these well documented cases, infection as a predisposing cause of postoperative pulmonary embolism was found to have an incidence equal to heart disease and obesity. Infection in patients with heart disease or obesity increased the

Postoperative Pulmonary Embolism

overall rate of pulmonary embolism associated with infection to 39 per cent. Infection, the factor most closely associated with pulmonary embolism fol- lowing abdominal or pelvic operation, occurred in eleven of seventeen patients (65 per cent). (Table II.) As a predisposing factor, infection after injuries of the lower extremities and amputation was equivalent to heart disease and obesity. The infection rate in patients with fatal postoperative pulmonary embolism was 48 per cent, whereas a 30 per cent rate of infection was noted in patients with nonfatal postoperative pulmonary embolism. The overall infection rate was 7 per cent on our surgical service, with a clean wound infection rate of 2.5 per cent.

Comments

The actual incidence of significant pulmonary emboli is obscured by difficulties in diagnosis and failure to pursue available diagnostic technics in cases of suspected pulmonary embolism. Autopsy series, which usually include both medical and surgical patients, cite overall incidences of pulmonary embolism as 2 per cent [IO] to 60 per cent [II]. The incidence of fatal pulmonary embolism diagnosed clinically was 0.03 to 0.7 per cent of medical, surgical, and obstetric cases [12]. Nonfatal postoperative pulmonary embolism was estimated by Coon and Willis [13] in 1959 to be three times more common than fatal postoperative pulmonary embolism. Sev- eral investigators [3,4] have reported fatal pulmonary embolism in approximately 1 per cent of postoperative patients. Our data indicate that 0.41 per cent of postoperative patients experience either a fatal or clinically documented nonfatal pulmonary embolus. Failure to perform autopsy on all patients who die, failure to make accurate clinical diagnoses, and inability to document whether emboli disclosed at autopsy caused death confound efforts to arrive at a reasonable estimate of the incidence of postoperative emboli. In our investigation, clinically significant nonfatal or fatal pulmonary emboli were sought,

TABLE II Risk Factors Associated with Pulmonary Embolism

Orthopedic Surgery, Abdominal or Pelvic Lower Extremity Injury,

No. of Operation (n = 17) Amputation (n = 19)

Incidence of No. of Incidence of

Infection Heart disease Obesity Aae >60 vr

Patient5 Association

11 65% 4 24% 5 29% 5 29%

Patients Association

4 21% 5 26% 7 37% 8 42%

Volume 137, March 1979 359

West, Garrison, and Flint

and thus the overall incidence is likely to be under- estimated.

Infection was found to be an equivalent risk factor to obesity and heart disease. Notable exception to this was found in the group of patients having abdominal or pelvic surgery, in which 65 per cent of those having pulmonary emboli had clinical infection. Infection was by far the most significant factor in the development of pulmonary embolism among patients who had abdominal operations. The exact relationship between infection and pulmonary embolism is not well understood. Infected patients are subject to prolonged bed rest and venous stasis. In addition, associated chronic diseases, particularly obesity and heart disease, are found with increasing frequency in patients undergoing abdominal operations [14]. Pelvic and abdominal operations may directly injure veins and thus facilitate the development of thrombosis.

Alterations in coagulation are associated with ex- perimental and clinical infections. Direct endothelial damage [15] and disseminated intravascular coagulation accompany septic shock [16]. Trauma, peri- tonitis, and intraabdominal abscess alter pulmonary hemodynamics [17], thus lessening the ability of the lungs to lyse emboli [18]. Bacterial endotoxin acti- vates factor IX, producing a hypercoaguable state and thrombosis in the area of venous stasis [19,20]. Other experiments have disclosed changes in levels of platelets [21] and fibrogen [22] in septic patients. The degree of infection necessary to change vessels or produce coagulopathy is unknown. Few of the infected patients in this series had septic shock, al- though all were febrile and had definite foci at infection.

Altemeier, Hill, and Fullen [23] in 1969 reported on the association of bacterial “L” type forms with acute and recurrent thromboembolic disease. Doc- umented sources of infection were found among some of their patients with thrombophlebitis. Their data suggested that the bacterial “L” type forms, particularly anaerobic bacteroides, play an etiologic role in thrombophlebitis and pulmonary embolus. One group of patients in that study was similar to patients in our series, in that both groups had a single attack of thrombophlebitis and pulmonary embolism after operation.

Our data disclose the important relationship between postoperative infection and pulmonary embolus in patients having abdominal and pelvic operations. Prophylactic minidose heparinization should be considered for patients undergoing general surgical procedures associated with an increased incidence of wound infection or deep abscess. Our

current practice for prophylactic minidose heparin is the subcutaneous administration of 5,000 units 2 hours prior to operation and 5,000 units every 12 hours thereafter.

Summary

Forty-one patients with significant pulmonary emboli were reviewed. The presence of heart disease and obesity was found with relatively constant frequency throughout the groups presented. Postop- erative infections (wound or abscess) were frequently associated with fatal or significant pulmonary embolism, particularly after abdominal and pelvic operations. The infection rate (65 per cent) in patients in whom significant pulmonary embolism developed after abdominal and pelvic operations is particularly striking when compared to the overall infection rate of 7 per cent for major operations in our hospital. Recent studies of prophylactic minidose heparinization reveal an increased number of complications due to the heparin [24], and thus the proper selection of cases for prophylactic minidose heparin is man- datory. The data suggest that patients at high risk for the development of postoperative complications of infection are also at high risk for the development of significant pulmonary emboli and should be considered candidates for prophylactic minidose heparinization.

References

1. Virchow R: Die verstopfund der iungenarterir und ihre folgen. Beitr Exp Pathol Physiol2: 1, 1946.

2. Sevitt S, Gallagher NG: Prevention of venous thrombosis and pulmonary embolism in injured patients. Lancet 2: 981, 1959.

3. Gjores JE: The incidence of venous thrombosis and its sequelae in certain districts of Sweden. i\cta Chir Stand [ Suppl] 206, 1956.

4. DeBakey ME: A critical evaluation of the problem of thromboembolism. ht Abs Surg 98: 1, 1954.

5. Kakkar VV, Corrigan TP, Fossard DP: Prevention of fatal postoperative pulmonary embolism by low doses of heparin. Lancet 2: 45, 1975.

6. Salzman EW, Harris WH, DeSanctis RW: Reduction in venous thromboembolism by agents affecting platelet function. N Engl J Med 284: 1287, 1971.

7. Evarts CM, Feil FJ: Prevention of thromboembolic disease after elective surgery of the hip. J Bone Joint Surg 53: 1271, 1971.

8. Williams HT: Prevention of postoperative deep-vein thrombosis with perioperative SQ heparin. Lancet 2: 950, 1971.

9. Gallus AS, Hirsh J, Tuttle RJ, Trebilcock R, O’Brien SE, Carroll JJ, Minden JH, Hudeck SM: Small subcutaneous doses of heparin in prevention of venous thrombosis. N Engl J Med 228: 545, 1973.

10. Pulvertaft RJV: Postoperative pulmonary embolism. Ann R Co/l Surg 1: 181, 1947.

11. Frieman DG, Suyemota J, Wessler S: Frequency of pulmonary thromboembolism in man. N Engl J Med 272: 1276, 1965.

12. Hume M, Sevitt S, Thomas DP: Venous Thrombosis and Pul-

360 The American Jogrnal ol Surgprv

Postoperative Pulmonary Embolism

monary Embolism. Cambridge, Harvard University Press, 1970.

13. Coon WW, Willis PW: Deep venous thrombosis and pulmonary embolism: prediction, prevention and treatment. Am,/ Car- diol4: 611, 1959.

14. Baughman BB. Polk HC Jr, Ahmad W, Knutson CO: The surgical treatment of carcinoma of colon and rectum. Ann Surg 183: 550, 1976.

15. Muller-Berghaus G, Lasch HG: Consumption coagulopathy in the course of endotoxinemia. Gram-Negative Bacterial In- fections and Mode of Endotoxin Actions (Urbaschek B, Ur- baschek R, Neter E, eds). Austria, Springer-VerlaglWien, 1975.

16. Gaynor E, Bouvier C, Spaet TH: Vascular lesion: possible pathogenic basis for the generalized Shwartzman reaction. Science 170: 986, 1970.

17. Harrison LH Jr, Hinshaw LB, Coalson JJ, Greenfield LF: Effects of E. coli and septic shock on pulmonary hemodynamics and capillary permeability. J Thorac Cardiovasc Surg 61: 795, 1971.

18. Rothman DL, Schwartz SI: Effects of trauma on pulmonary plasminogen activator activity. Surg Forum 16: 94, 1975.

19. Thomas DP, Wessler S, Reiner SM: The relation of factors XII, XI and IX to hypercoaguable states. Thorn& Haemostas 9: 90, 1963.

20. Thomas DP, Wessler S: Stasis thrombi induced by bacterial endotoxin. Circ Res 14: 486, 1964.

21. Cowan DH, Bowman LS, Fratianne RB, Ahmed F: Platelet aggregation as a sign of septicemia in thermal injury. JAMA 235: 1230, 1976.

22. Corrigan J Jr, Ray WL. May N: Changes in the blood coagulation system associated with septicemia. N fngl J Med 279: 85 1, 1968.

23. Altemeier WA, Hill EO, Fullen WD: Acute and recurrent thromboembolic disease: a new concept in etiology. Ann Surg 170: 547, 1969.

24. Pachter HL, Riles TS: Low dose heparin: bleeding and wound complications in the surgical patient. A prospective ran- domized study. Ann Surg 186: 669, 1977.

Volunrq 137, March 1979 361

effects of concurrent sepsis with clinically significant pulmonary embolic disease

Documents