- Email: [email protected]
Role of antral puncture in the treatment of sinusitis in the intensive care unit
Role of antral puncture in the treatment of sinusitis in the intensive care unit HASSAN H. RAMADAN, MD, FACS, ROBERT M. OWENS, MD, CHRISTOPHER TIU, MD, and MARK K. WAX, MD, FACS,
Morgantown, West Virginia, and Buffalo, New York
OBJECTIVE: The objective of this study was to determine whether maxillary sinus puncture caused an alteration in antibiotic treatment and thus affected the outcome of sinusitis in the intensive care unit. STUDY DESIGN: A retrospective review was done of cases of maxillary sinus puncture between 1991 and 1994. RESULTS: Forty-two patients were identified. All patients had findings suggestive of sinusitis on plain sinus films or CT scans of the sinuses. Twenty-five punctures (60%) recovered pus, of which 80% grew organisms. Fourteen patients (33%) had a negative lavage, and 12% of these samples grew organisms (p = 0.001). Sixty-eight percent of the cultures identified a single organism compared with 32% with multiple organisms. Gram-negative organisms were found most commonly, followed by anaerobes. In 57% of the cases antibiotic therapy was changed. In 77% of the cases the change was directed by the culture result and in 35% the therapy was changed despite a negative culture result (p = 0.002). Resolution of symptoms occurred in 83% of patients who had antibiotics changed whereas with no change in antibiotics only 42% had resolution (p = 0.001). CONCLUSION: A sinus puncture seems to be helpful in patients with fever and positive findings on sinus films. If pus is obtained by maxillary puncture, a positive culture may be found in 80% of the cases. Changes in antibiotic regimen on the basis of culture findings seem to give a better outcome. (Otolaryngol Head Neck Surg 1998;119:381-4.)
From the Department of Otolaryngology–Head and Neck Surgery, West Virginia University, Morgantown (Drs. Ramadan, Owens, and Tiu), and Buffalo General Hospital, Buffalo, New York (Dr. Wax). Presented at the Annual Meeting of the American Academy of Otolaryngology–Head and Neck Surgery, Washington, D.C., Sept. 29–Oct. 2, 1996. Reprint requests: Hassan H. Ramadan, MD, FACS, Department of Otolaryngology–Head and Neck Surgery, West Virginia University, P.O. Box 9200, Morgantown WV 26506-9200. Copyright © 1998 by the American Academy of Otolaryngology– Head and Neck Surgery Foundation, Inc. 0194-5998/98/$5.00 + 0 23/1/84678
Paranasal
sinusitis is a common condition that is encountered by medical personnel caring for patients in the intensive care unit (ICU). It has also been recognized as an important nosocomial infection in the ICU. The difficulty in diagnosing sinusitis in the ICU lies in that the infection is often clinically silent. The majority of the patients are free of symptoms and sinus abnormalities are detected on imaging studies done for reasons other than evaluating the sinuses. Otolaryngologists are frequently consulted to evaluate such conditions and to give recommendations about performing an antral puncture. The literature, however, is unclear regarding how antral puncture findings affect eventual patient outcome. These patients are usually critically ill and the ICU personnel are looking for a source of infection to direct treatment of these patients. The sequelae can be catastrophic and early intervention and diagnostic inquiry are justifiable. However, before we ask ourselves whether antral puncture is helpful, it seems reasonable initially to determine whether antral puncture has had any impact on the treatment of these patients. The purpose of this study was to evaluate the role of antral puncture in the treatment of sinusitis in the ICU and, specifically, how alterations in antibiotic treatment were initiated on the basis of puncture results and how this affected patient outcomes. METHODS AND MATERIAL The charts of all patients who had an antral puncture done between 1991 and 1994 were retrospectively reviewed. The diagnosis of sinusitis was based on radiologic and clinical findings. A total of 42 patients with complete records were identified (Table 1). Twenty-three patients were in the surgical ICU (55%) and 19 were in the medical ICU (45%). The average hospital stay was 12.5 days with an average febrile duration of 6.4 days. The mean age was 48.9 years with a range of 19 to 83 years. There were 28 men and 14 women. In the majority of the cases the admission diagnosis was trauma (18), and this was followed by sepsis (8), operation (5), central nervous system condition (5), and other conditions (6). None of the 42 patients had any nasal packing at the time of diagnosis. Seventeen patients had associated surgical or traumatic wounds. Ten patients had diabetes mellitus, 11 were tobacco and alcohol consumers, and 6 were receiving steroid treatment. In the remaining patients the medical histo381
Otolaryngology– Head and Neck Surgery October 1998
382 RAMADAN et al.
Fig. 1. Number of coexisting infections in all patients. CNS, Central nervous system.
Table 1. Demographic features of study population (n = 42) No. (%)
Male sex Unit Surgical Medical Admission diagnosis Trauma Sepsis Postop. Stroke Other Nasogastric tube Age (yr) Mean Range Pretap antibiotics (days) Mean Range
28 (67) 23 (55) 19 (45) 18 8 5 5 6 34 (81) 48.9 19-83 8.2 0-28
ry was unremarkable. Thirty-four patients had a nasogastric tube at the time of diagnosis. Coexistent infections consisted of pneumonia in 16, sepsis in 10, urinary tract infections in 6, wound infections in 4, meningitis in 2, throat infections in 3, abdominal infections in 3, and tracheotomy in 2; 10 patients had no associated infections (Fig. 1). Forty patients had been receiving antibiotic treatment before antral puncture for a mean of 8.2 days (range 0 to 28 days). All patients had radiologic evidence of sinusitis on a CT scan of the sinuses (22), a CT scan of the head (11), or plain sinus films (9). All patients had either an air fluid level in one or more sinuses or opacification of one or more sinuses. Sinusitis was suspected in these patients when the opacification or the air fluid level persisted despite antimicrobial treatment and removal of indwelling nasal tubes. Antral punctures were done on the maxillary sinus after injection of a local anesthetic to the canine fossa. A trocar was introduced into the maxillary sinus and any pus that was encountered was sent for aerobic and anaerobic culturing. If the tap failed to demonstrate pus, lavage of the sinus with saline solution was done and the fluid sent for culturing. Our outcome measure was res-
olution of the fever, and the independent variable was change in management, which was mainly a change in the antibiotics administered. Because any change of antibiotics was based on culture result, which depended on the antral puncture findings, analysis was done to determine whether a change in the antibiotic caused resolution of symptoms. For statistical analysis, χ2, Fisher’s exact test, t test, and multivariate logistic regression analysis were used. RESULTS
In 25 patients (60%) pus was recovered from the sinuses. Fourteen patients had a negative tap result and a bloody tap was encountered in the remaining three patients. Among the purulent tap results, 20 samples (80%) grew bacterial organisms whereas among the negative tap results only 2 samples (12%) grew bacterial organisms (p = 0.001). Antibiotics were changed in 77% of patients who had a positive culture compared with only 35% of those who had a negative culture. Antibiotics were not changed in five (23%) of the patients despite a positive culture. Resolution of fever occurred in 83% of the patients who had the antibiotics changed within that period, whereas only 42% had resolution of fever among those who did not have the antibiotics changed. Eleven patients died, nine of whom did not have resolution of fever; the remaining two had resolution of the fever before death. Univariate analysis of all the other factors was done and the results are summarized in Table 2. None of the variables was a predictor of the resolution of symptoms except for the duration of antibiotic treatment that patients had received before the tap. Multivariate Analysis
Logistic regression analysis was then done using all the variables to check for confounders. Our outcome measure was the resolution of the fever and the independent variable was the antibiotic change. Because antibiotic change was dependent on culture result, which was highly correlated with puncture findings,
Otolaryngology– Head and Neck Surgery Volume 119 Number 4
RAMADAN et al. 383
Fig. 2. Result of cultures obtained from sinus aspirates.
puncture findings and culture result were not included in the analysis. Age was used as a continuous variable in the analysis, but the duration of the antibiotics that patients received before the tap was dichotomized into two groups with the cut-off point set at 1 week. The analysis did not reveal any of the factors to be a confounder except for the duration of the pretap antibiotics. Patients who had received antibiotics for less than 1 week benefited the most from the change in antibiotics (p = 0.03). However, despite this finding, the independent variable of the change in antibiotics remained the sole predictor of outcome. Seventeen patients had growth of only one organism compared with eight who had multiple organisms (Fig. 2). The organisms seen were similar to those reported in the literature, namely nosocomial organisms. Gram-negative organisms and staphylococcus species were the most common organisms cultured. Complications of the tap included bleeding in three patients and a cardiac arrest at the time of the tap in one patient who was resuscitated successfully. No complications were encountered in the remaining 38 patients. DISCUSSION
Sinusitis in ICU patients is a relatively common condition. If not treated it can lead to serious morbidity and mortality. Because the majority of these patients are free of symptoms, sinusitis is usually detected during routine work-up of unresolving fever. Most of these patients have associated infections and are being treated with antibiotics.1 When sinusitis is detected on radiologic evaluation, the otolaryngologist is called on to do an antral puncture. Antral puncture is an important adjunct to treatment of sinusitis in these patients, because it serves to provide diagnostic information regarding microbiologic types. From a review of the literature, it is unclear how the diagnostic information provided by this invasive procedure is used and how it affects eventual patient outcomes.2 There seems to be
Table 2. Analysis of variables and the resolution of symptoms Variable
Sex Male Female Admission diagnosis Surgical Medical Infection Present Absent Wounds Present Absent Nasogastric tube Present Absent Puncture findings Positive Negative Unit Surgical Medical Culture Positive Negative Antibiotic change Yes No Mean age (yr) Mean antibiotic duration (days)
Resolved (%)
Nonresolved (%)
74 26
53 47
73 27
50 50
81 19
85 15
48 52
27 73
78 22
93 7
67 33
53 47
59 41
47 53
59 41
40 60
20 30 46.1 6.3
27 73 53.9 11.5
p Value
0.2 0.2 0.8 0.2 0.4 0.2 0.4 0.2 0.006 0.25 0.02
an agreement, however, that prolonged use of nasal tubes predisposes patients in the ICU to the development of sinusitis.3-5 During the past two decades there has been an increasing interest in the incidence of sinusitis in ICU patients. In 1974, Arens et al.4 reported on cases of sinusitis in ICU patients who had been nasotracheally intubated and suggested that the mechanical obstruction of sinus ostia by nasotracheal tubes predisposed patients to the development of sinusitis. Kronberg and
Otolaryngology– Head and Neck Surgery October 1998
384 RAMADAN et al.
Goodwin,6 in 1985, reported on six cases of sinusitis in ICU patients. The authors speculated that in five patients sepsis developed that was related to sinusitis. Predisposing factors described included prolonged indwelling nasal tubes and immunosuppression via corticosteroids or diabetes. Humphrey et al.5 in 1987 reported on the development of sinusitis in 24 of 208 patients with head injuries. This article emphasized the role that sinusitis may play in the development of infections in other anatomic locations, in that many patients had the development of tracheobronchitis, pneumonia, bacteremia, and meningitis with organisms identical to those that had been recovered from sinus aspirations. Reports published by Bell et al.7 in 1988 and Linden et al.3 in 1988 also concluded that sinusitis is a relatively frequent finding in ICU patients and may be a significant cause of morbidity and mortality in this patient population. Most recently, Borman et al.1 in 1992 reported a series of 26 patients among 598 admissions to the ICU. The authors again stressed that the majority of patients had prolonged indwelling nasal tubes and imaging studies that demonstrated sinus abnormalities. The authors also pointed out that sinusitis is seldom the sole source of fever in this patient population because of the high coexistence of other infections. From a review of the relevant literature, it is apparent that ICU patients are predisposed to sinusitis and that there is a relatively high incidence of this finding in this patient population. Predisposing factors include indwelling nasal tubes, immunosuppression, and probably positioning of the patients. ICU sinusitis has been linked to concurrent infections and cited as an infection sequela. The majority of infections are found to arise from nosocomial organisms and are multiorganism as opposed to single organism infections. ICU patients do frequently have coexistent infections and it is difficult to determine with certainty how often sinusitis represents the sole cause of fever in this patient population. Univariate analysis of the data showed that in 60% of the cases, pus was recovered during the tap. When pus was recovered, 80% of the samples grew microorganisms on cultures, but only 12% of the samples showed microorganisms when the tap was negative (p = 0.001). Antibiotics were changed in 77% of the cases with positive culture findings compared with only 35% of those with negative culture findings (p = 0.002). When the antibiotics were changed, 83% of the patients had resolution of fever. Only 42% of patients had resolution of fever when the antibiotics were not changed (p = 0.006).
Analysis of the other variables did not reveal any significant impact on the resolution of symptoms except for the duration of antibiotics that the patients received before the tap. In those who had resolution of fever, the mean duration was 6.3 days compared with 11.5 days for those who had no resolution of symptoms (p = 0.02). Multivariate analysis however, revealed that the change of antibiotics was the only significant factor predicting resolution of the fever (p < 0.05). CONCLUSION
Antral puncture in ICU patients with suspected sinusitis was helpful in guiding ICU staff on which antibiotics to use. When antibiotics were changed accordingly, resolution of the fever and the sinusitis occurred in 83% of the cases compared with resolution in only 42% when antibiotics were not changed. Despite the fact that therapeutic changes initiated by findings from antral puncture appear to improve resolution of fever, we cannot conclude that fever or sepsis arose solely from sinusitis, inasmuch as it is well established that the majority of these ICU patients may have coexistent infections. With regard to comparing medical management alone versus management based on antral punctures findings, a randomized, blinded prospective study would be necessary. Given the potentially hazardous sequelae of not aggressively treating a sinus infection that is suspected as a source of sepsis in these patients, this type of prospective study may not be ethical to perform. We thank Kimberly ReVeal and Beth Ann Dlugos for their assistance in helping to prepare this manuscript. REFERENCES 1. Borman KR, Brown PM, Mezera KK, Jhaveri H. Occult fever in surgical intensive care unit patients is seldom caused by sinusitis. Am J Surg 1992;164:412-6. 2. Lebeda MD, Haller JR, Graham SM, Hoffman HT. Evaluation of maxillary sinus aspiration in patients with fever of unknown origin. Laryngoscope 1995;105:683-5. 3. Linden BE, Aguilar EA, Allen SJ. Sinusitis in the nasotracheally intubated patient. Arch Otolaryngol Head Neck Surg 1988; 114:860-1. 4. Arens JF, LeJeune FE Jr, Webre DR. Maxillary sinusitis, a complication of nasotracheal intubation. Anesthesiology 1974;40: 415-6. 5. Humphrey MA, Simpson GT, Grindlinger GA. Clinical characteristics of nosocomial sinusitis. Ann Otol Rhinol Laryngol 1987;96:687-90. 6. Kronberg FG, Goodwin WJ Jr. Sinusitis in intensive care unit patients. Laryngoscope 1985;95:936-8. 7. Bell RM, Page GV, Bynoe RP, Dunham ME, Brill AH. Posttraumatic sinusitis. J Trauma 1988;28:923-30.
Morgantown, West Virginia, and Buffalo, New York
OBJECTIVE: The objective of this study was to determine whether maxillary sinus puncture caused an alteration in antibiotic treatment and thus affected the outcome of sinusitis in the intensive care unit. STUDY DESIGN: A retrospective review was done of cases of maxillary sinus puncture between 1991 and 1994. RESULTS: Forty-two patients were identified. All patients had findings suggestive of sinusitis on plain sinus films or CT scans of the sinuses. Twenty-five punctures (60%) recovered pus, of which 80% grew organisms. Fourteen patients (33%) had a negative lavage, and 12% of these samples grew organisms (p = 0.001). Sixty-eight percent of the cultures identified a single organism compared with 32% with multiple organisms. Gram-negative organisms were found most commonly, followed by anaerobes. In 57% of the cases antibiotic therapy was changed. In 77% of the cases the change was directed by the culture result and in 35% the therapy was changed despite a negative culture result (p = 0.002). Resolution of symptoms occurred in 83% of patients who had antibiotics changed whereas with no change in antibiotics only 42% had resolution (p = 0.001). CONCLUSION: A sinus puncture seems to be helpful in patients with fever and positive findings on sinus films. If pus is obtained by maxillary puncture, a positive culture may be found in 80% of the cases. Changes in antibiotic regimen on the basis of culture findings seem to give a better outcome. (Otolaryngol Head Neck Surg 1998;119:381-4.)
From the Department of Otolaryngology–Head and Neck Surgery, West Virginia University, Morgantown (Drs. Ramadan, Owens, and Tiu), and Buffalo General Hospital, Buffalo, New York (Dr. Wax). Presented at the Annual Meeting of the American Academy of Otolaryngology–Head and Neck Surgery, Washington, D.C., Sept. 29–Oct. 2, 1996. Reprint requests: Hassan H. Ramadan, MD, FACS, Department of Otolaryngology–Head and Neck Surgery, West Virginia University, P.O. Box 9200, Morgantown WV 26506-9200. Copyright © 1998 by the American Academy of Otolaryngology– Head and Neck Surgery Foundation, Inc. 0194-5998/98/$5.00 + 0 23/1/84678
Paranasal
sinusitis is a common condition that is encountered by medical personnel caring for patients in the intensive care unit (ICU). It has also been recognized as an important nosocomial infection in the ICU. The difficulty in diagnosing sinusitis in the ICU lies in that the infection is often clinically silent. The majority of the patients are free of symptoms and sinus abnormalities are detected on imaging studies done for reasons other than evaluating the sinuses. Otolaryngologists are frequently consulted to evaluate such conditions and to give recommendations about performing an antral puncture. The literature, however, is unclear regarding how antral puncture findings affect eventual patient outcome. These patients are usually critically ill and the ICU personnel are looking for a source of infection to direct treatment of these patients. The sequelae can be catastrophic and early intervention and diagnostic inquiry are justifiable. However, before we ask ourselves whether antral puncture is helpful, it seems reasonable initially to determine whether antral puncture has had any impact on the treatment of these patients. The purpose of this study was to evaluate the role of antral puncture in the treatment of sinusitis in the ICU and, specifically, how alterations in antibiotic treatment were initiated on the basis of puncture results and how this affected patient outcomes. METHODS AND MATERIAL The charts of all patients who had an antral puncture done between 1991 and 1994 were retrospectively reviewed. The diagnosis of sinusitis was based on radiologic and clinical findings. A total of 42 patients with complete records were identified (Table 1). Twenty-three patients were in the surgical ICU (55%) and 19 were in the medical ICU (45%). The average hospital stay was 12.5 days with an average febrile duration of 6.4 days. The mean age was 48.9 years with a range of 19 to 83 years. There were 28 men and 14 women. In the majority of the cases the admission diagnosis was trauma (18), and this was followed by sepsis (8), operation (5), central nervous system condition (5), and other conditions (6). None of the 42 patients had any nasal packing at the time of diagnosis. Seventeen patients had associated surgical or traumatic wounds. Ten patients had diabetes mellitus, 11 were tobacco and alcohol consumers, and 6 were receiving steroid treatment. In the remaining patients the medical histo381
Otolaryngology– Head and Neck Surgery October 1998
382 RAMADAN et al.
Fig. 1. Number of coexisting infections in all patients. CNS, Central nervous system.
Table 1. Demographic features of study population (n = 42) No. (%)
Male sex Unit Surgical Medical Admission diagnosis Trauma Sepsis Postop. Stroke Other Nasogastric tube Age (yr) Mean Range Pretap antibiotics (days) Mean Range
28 (67) 23 (55) 19 (45) 18 8 5 5 6 34 (81) 48.9 19-83 8.2 0-28
ry was unremarkable. Thirty-four patients had a nasogastric tube at the time of diagnosis. Coexistent infections consisted of pneumonia in 16, sepsis in 10, urinary tract infections in 6, wound infections in 4, meningitis in 2, throat infections in 3, abdominal infections in 3, and tracheotomy in 2; 10 patients had no associated infections (Fig. 1). Forty patients had been receiving antibiotic treatment before antral puncture for a mean of 8.2 days (range 0 to 28 days). All patients had radiologic evidence of sinusitis on a CT scan of the sinuses (22), a CT scan of the head (11), or plain sinus films (9). All patients had either an air fluid level in one or more sinuses or opacification of one or more sinuses. Sinusitis was suspected in these patients when the opacification or the air fluid level persisted despite antimicrobial treatment and removal of indwelling nasal tubes. Antral punctures were done on the maxillary sinus after injection of a local anesthetic to the canine fossa. A trocar was introduced into the maxillary sinus and any pus that was encountered was sent for aerobic and anaerobic culturing. If the tap failed to demonstrate pus, lavage of the sinus with saline solution was done and the fluid sent for culturing. Our outcome measure was res-
olution of the fever, and the independent variable was change in management, which was mainly a change in the antibiotics administered. Because any change of antibiotics was based on culture result, which depended on the antral puncture findings, analysis was done to determine whether a change in the antibiotic caused resolution of symptoms. For statistical analysis, χ2, Fisher’s exact test, t test, and multivariate logistic regression analysis were used. RESULTS
In 25 patients (60%) pus was recovered from the sinuses. Fourteen patients had a negative tap result and a bloody tap was encountered in the remaining three patients. Among the purulent tap results, 20 samples (80%) grew bacterial organisms whereas among the negative tap results only 2 samples (12%) grew bacterial organisms (p = 0.001). Antibiotics were changed in 77% of patients who had a positive culture compared with only 35% of those who had a negative culture. Antibiotics were not changed in five (23%) of the patients despite a positive culture. Resolution of fever occurred in 83% of the patients who had the antibiotics changed within that period, whereas only 42% had resolution of fever among those who did not have the antibiotics changed. Eleven patients died, nine of whom did not have resolution of fever; the remaining two had resolution of the fever before death. Univariate analysis of all the other factors was done and the results are summarized in Table 2. None of the variables was a predictor of the resolution of symptoms except for the duration of antibiotic treatment that patients had received before the tap. Multivariate Analysis
Logistic regression analysis was then done using all the variables to check for confounders. Our outcome measure was the resolution of the fever and the independent variable was the antibiotic change. Because antibiotic change was dependent on culture result, which was highly correlated with puncture findings,
Otolaryngology– Head and Neck Surgery Volume 119 Number 4
RAMADAN et al. 383
Fig. 2. Result of cultures obtained from sinus aspirates.
puncture findings and culture result were not included in the analysis. Age was used as a continuous variable in the analysis, but the duration of the antibiotics that patients received before the tap was dichotomized into two groups with the cut-off point set at 1 week. The analysis did not reveal any of the factors to be a confounder except for the duration of the pretap antibiotics. Patients who had received antibiotics for less than 1 week benefited the most from the change in antibiotics (p = 0.03). However, despite this finding, the independent variable of the change in antibiotics remained the sole predictor of outcome. Seventeen patients had growth of only one organism compared with eight who had multiple organisms (Fig. 2). The organisms seen were similar to those reported in the literature, namely nosocomial organisms. Gram-negative organisms and staphylococcus species were the most common organisms cultured. Complications of the tap included bleeding in three patients and a cardiac arrest at the time of the tap in one patient who was resuscitated successfully. No complications were encountered in the remaining 38 patients. DISCUSSION
Sinusitis in ICU patients is a relatively common condition. If not treated it can lead to serious morbidity and mortality. Because the majority of these patients are free of symptoms, sinusitis is usually detected during routine work-up of unresolving fever. Most of these patients have associated infections and are being treated with antibiotics.1 When sinusitis is detected on radiologic evaluation, the otolaryngologist is called on to do an antral puncture. Antral puncture is an important adjunct to treatment of sinusitis in these patients, because it serves to provide diagnostic information regarding microbiologic types. From a review of the literature, it is unclear how the diagnostic information provided by this invasive procedure is used and how it affects eventual patient outcomes.2 There seems to be
Table 2. Analysis of variables and the resolution of symptoms Variable
Sex Male Female Admission diagnosis Surgical Medical Infection Present Absent Wounds Present Absent Nasogastric tube Present Absent Puncture findings Positive Negative Unit Surgical Medical Culture Positive Negative Antibiotic change Yes No Mean age (yr) Mean antibiotic duration (days)
Resolved (%)
Nonresolved (%)
74 26
53 47
73 27
50 50
81 19
85 15
48 52
27 73
78 22
93 7
67 33
53 47
59 41
47 53
59 41
40 60
20 30 46.1 6.3
27 73 53.9 11.5
p Value
0.2 0.2 0.8 0.2 0.4 0.2 0.4 0.2 0.006 0.25 0.02
an agreement, however, that prolonged use of nasal tubes predisposes patients in the ICU to the development of sinusitis.3-5 During the past two decades there has been an increasing interest in the incidence of sinusitis in ICU patients. In 1974, Arens et al.4 reported on cases of sinusitis in ICU patients who had been nasotracheally intubated and suggested that the mechanical obstruction of sinus ostia by nasotracheal tubes predisposed patients to the development of sinusitis. Kronberg and
Otolaryngology– Head and Neck Surgery October 1998
384 RAMADAN et al.
Goodwin,6 in 1985, reported on six cases of sinusitis in ICU patients. The authors speculated that in five patients sepsis developed that was related to sinusitis. Predisposing factors described included prolonged indwelling nasal tubes and immunosuppression via corticosteroids or diabetes. Humphrey et al.5 in 1987 reported on the development of sinusitis in 24 of 208 patients with head injuries. This article emphasized the role that sinusitis may play in the development of infections in other anatomic locations, in that many patients had the development of tracheobronchitis, pneumonia, bacteremia, and meningitis with organisms identical to those that had been recovered from sinus aspirations. Reports published by Bell et al.7 in 1988 and Linden et al.3 in 1988 also concluded that sinusitis is a relatively frequent finding in ICU patients and may be a significant cause of morbidity and mortality in this patient population. Most recently, Borman et al.1 in 1992 reported a series of 26 patients among 598 admissions to the ICU. The authors again stressed that the majority of patients had prolonged indwelling nasal tubes and imaging studies that demonstrated sinus abnormalities. The authors also pointed out that sinusitis is seldom the sole source of fever in this patient population because of the high coexistence of other infections. From a review of the relevant literature, it is apparent that ICU patients are predisposed to sinusitis and that there is a relatively high incidence of this finding in this patient population. Predisposing factors include indwelling nasal tubes, immunosuppression, and probably positioning of the patients. ICU sinusitis has been linked to concurrent infections and cited as an infection sequela. The majority of infections are found to arise from nosocomial organisms and are multiorganism as opposed to single organism infections. ICU patients do frequently have coexistent infections and it is difficult to determine with certainty how often sinusitis represents the sole cause of fever in this patient population. Univariate analysis of the data showed that in 60% of the cases, pus was recovered during the tap. When pus was recovered, 80% of the samples grew microorganisms on cultures, but only 12% of the samples showed microorganisms when the tap was negative (p = 0.001). Antibiotics were changed in 77% of the cases with positive culture findings compared with only 35% of those with negative culture findings (p = 0.002). When the antibiotics were changed, 83% of the patients had resolution of fever. Only 42% of patients had resolution of fever when the antibiotics were not changed (p = 0.006).
Analysis of the other variables did not reveal any significant impact on the resolution of symptoms except for the duration of antibiotics that the patients received before the tap. In those who had resolution of fever, the mean duration was 6.3 days compared with 11.5 days for those who had no resolution of symptoms (p = 0.02). Multivariate analysis however, revealed that the change of antibiotics was the only significant factor predicting resolution of the fever (p < 0.05). CONCLUSION
Antral puncture in ICU patients with suspected sinusitis was helpful in guiding ICU staff on which antibiotics to use. When antibiotics were changed accordingly, resolution of the fever and the sinusitis occurred in 83% of the cases compared with resolution in only 42% when antibiotics were not changed. Despite the fact that therapeutic changes initiated by findings from antral puncture appear to improve resolution of fever, we cannot conclude that fever or sepsis arose solely from sinusitis, inasmuch as it is well established that the majority of these ICU patients may have coexistent infections. With regard to comparing medical management alone versus management based on antral punctures findings, a randomized, blinded prospective study would be necessary. Given the potentially hazardous sequelae of not aggressively treating a sinus infection that is suspected as a source of sepsis in these patients, this type of prospective study may not be ethical to perform. We thank Kimberly ReVeal and Beth Ann Dlugos for their assistance in helping to prepare this manuscript. REFERENCES 1. Borman KR, Brown PM, Mezera KK, Jhaveri H. Occult fever in surgical intensive care unit patients is seldom caused by sinusitis. Am J Surg 1992;164:412-6. 2. Lebeda MD, Haller JR, Graham SM, Hoffman HT. Evaluation of maxillary sinus aspiration in patients with fever of unknown origin. Laryngoscope 1995;105:683-5. 3. Linden BE, Aguilar EA, Allen SJ. Sinusitis in the nasotracheally intubated patient. Arch Otolaryngol Head Neck Surg 1988; 114:860-1. 4. Arens JF, LeJeune FE Jr, Webre DR. Maxillary sinusitis, a complication of nasotracheal intubation. Anesthesiology 1974;40: 415-6. 5. Humphrey MA, Simpson GT, Grindlinger GA. Clinical characteristics of nosocomial sinusitis. Ann Otol Rhinol Laryngol 1987;96:687-90. 6. Kronberg FG, Goodwin WJ Jr. Sinusitis in intensive care unit patients. Laryngoscope 1985;95:936-8. 7. Bell RM, Page GV, Bynoe RP, Dunham ME, Brill AH. Posttraumatic sinusitis. J Trauma 1988;28:923-30.