150
Asymptomatic Bacteriuria in Patients with Diabetes Mellitus G. G. Zhanel, G. K. M. Harding, and L. E. Nicolle
From the Departments of Medicine and Medical Microbiology, Faculty of Medicine and the Faculty of Pharmacy, University of Manitoba; the Section of Infectious Diseases, St. Boniface General Hospital and Health Sciences Centre; and the St. Boniface General Hospital Research Institute, Winnipeg, Manitoba, Canada
The urinary tract is a frequent site of infection in diabetic patients [1]. Data derived from postmortem evaluationin the preantibiotic era document a histologic prevalenceof "pyelonephritis" of rv 20 % among patients with diabetes mellitus and the occurrence of serious renal infections five times as frequently in diabetic than in nondiabeticpatients [2,3]. Urinary tract infections not only occur more frequently but also are more protracted and have a more serious outcome in diabeticsthan in nondiabetics [3]. Studiesdo notagreeon whether the majority of pyelonephritis cases are the result of hematogenous dissemination or ascending infection, but ascending infectioncertainly contributesa substantialamount to this entity [4, 5]. Studies demonstrate greater susceptibility of diabetic than of nondiabetic animalsto bacteriuriaand pyelonephritis [6-8]. In addition, insulin regulation in diabetic rats does not influence the incidence of infection [6]. Suggested mechanisms for the greater susceptibility of the diabetic urinary tract to
Received for publication30 October 1989and in revised form 24 April 1990. Dr. Zhanel is the recipient of a Postdoctoral Research Fellowship in Medical Microbiology from the Pharmaceutical Manufacturers Association of Canada Health Research Foundation and the Medical Research Council of Canada. The authors thank L. Dingman and S. Singh for their expert secretarial assistance. Please address requests for reprints to Dr. G. G. Zhanel, Department of Microbiology, MS-6, Health Sciences Centre, 820 Sherbrook, Winnipeg, Manitoba, CanadaR3A lR9. Reviews of Infectious Diseases 1991;13:150-4 © 1991 by The University of Chicago. All rights reserved. 0162-0886/91/1301-0067$02.00
infection include decreased antibacterial activity of the urine as a result of dilution of inhibitory substances such as urea, defects in polymorphonuclear leukocyte function or cellular immunity, and increased adhesivecapacity of bladder epithelial cells [6-8]. Several factors may place patients with diabetes mellitus at risk for asymptomaticbacteriuria, including poor control of blood glucoselevels, diabetic neuropathy with neurogenic bladder, impairment of leukocyte function as a result of hyperglycemia, frequent instrumentation of the urinary tract, and diabetic microangiopathy [9, 10]. Asymptomaticbacteriuria is generally definedas the presence of ~108 of a single bacterial species/L (~I()5 cfu/mL) in at least two clean-voided specimens of midstreamurinefrom an individual with no symptoms referable to the genitourinary tract [11]. This review will consider several aspects of asymptomatic bacteriuria in patients with diabetes mellitus, including prevalence, bacteriology, associatedrisk factors, localizationto the upper tract (kidney)or lower tract (bladder), natural history, and treatment.
Prevalence Several studies haveassessed the prevalenceof asymptomatic bacteriuria in women with diabetes mellitus (table I) [12-22]. These studies report a prevalence ranging from 0 to 29.3 % in differentpopulationsof diabetic women. Among nondiabetic control womenthe prevalenceranges from 2.8 % to 22.0 %. The majority of investigatorshavereported a prevalenceofasymptomatic bacteriuriathat is approximately threefoldhigher among womenwith diabetes mellitus than among nondiabetic women [12-26].
Downloaded from http://cid.oxfordjournals.org/ at Fachhochschule Osnabrueck Zentrale Einrichtung fuer Wissenschaftliche Infor on April 16, 2015
This review analyzes several aspects of asymptomatic bacteriuria in patients with diabetes mellitus, including prevalence, bacteriology, coexistent risk factors, localization, natural history, and treatment. The prevalence ofasymptomatic bacteriuria is threefold higher among diabetic women than among nondiabetic women. However, rates among diabetic and nondiabetic men are similar. Studies consistently document that the prevalence of asymptomatic bacteriuria is not influenced by the type or duration of diabetes or by the quality of diabetic control. The microorganisms causing asymptomatic bacteriuria in persons with diabetes mellitus are similar to those causing bacteriuria in nondiabetic individuals. More than half of diabetic patients with asymptomatic bacteriuria have upper urinary tract involvement. The long-term consequences of asymptomatic bacteriuria in patients with diabetes mellitus are poorly documented. Clinical trials describing the treatment of asymptomatic bacteriuria in this population suggest that a 2-week course of therapy is equivalent to a 6-week course for the initial eradication of bacteriuria and that, following treatment, reinfection rather than relapse usually occurs. Important questions remaining include whether asymptomatic bacteriuria should be treated in this population and what the optimal antimicrobial regimen is.
RID 1991;13 (January-February)
Table 1.
Asymptomatic Bacteriuria in Diabetes
151
Prevalence of asymptomatic bacteriuria among women with diabetes mellitus. Diabetic women
Reference
Nondiabetic women
Type of population
Age range in y (mean)
Prevalence
81 128 20 40 152 111 230
70 (-45) 0->70 (31.1) 24-59 (42) 61-82 (68) 10-60 (-55)
341 41 92
Outpatient Outpatient Outpatient Outpatient Outpatient Outpatient Outpatient Outpatient Outpatient Hospitalized Outpatient
15-65 (46) 32-80 (55) 17-79 (>50) 0->60 (...)
15 (18.5) 24 (18.8) 0 8 (20.0) 24 (15.8) 30 (27.0) 43 (18.7) 38 (9.5) 31 (9.1) 12 (29.3) 18 (19.6)
100 41 92
100
Outpatient
10-69 (-40)
9 (9.0)
100
12 13
14 14 15 16 17 18 19 20 21 22
400
Adult" Adult"
(%)
No. of patients
Type of population
81 114 36
Age range in y (mean)
Prevalence
Outpatient Outpatient Outpatient
70 (-45) 10->70 (38) 61-88 (72)
3 (3.7) 9 (7.9) 1 (2.8)
Outpatient Outpatient
20-60 (-55)
(%)
· .. ( ... )
*
152 79
Adult"
7 (4.6) 9 (11.4)
· .. ( ... ) · .. ( ... ) Outpatient Hospitalized Casualty department Outpatient
· .. (54) 17-79 (>50) 0->60 (...)
5 (5.0) 9 (22.0) 17 (18.5)
10-69 (-40)
8 (8.0)
* Ellipses indicate that data are not available. t More specific data are not available.
Table 2.
Prevalence of asymptomatic bacteriuria among men with diabetes mellitus. Diabetic men
Reference
12 13 14 15 16 17 19 20 21 22
Nondiabetic men
No. of patients
Age range in y (mean)
Prevalence
67 141 40 154 87 103 411 9 58 90
70 (45) 0->70 (44.4) 16-77 (54) 20-60 (-52)
5 (7.5) 1 (0.7) 1 (2.5) 2 (1.3) 7 (8.0) 2 (1.9) 4 (1.0) 1(11.1) 2 (3.4) 3 (3.3)
Adultf Adultt
32-80 (55) 17-79 (>50) 0->60 (...) 10-69 (-40)
(%)
No. of patients
Age range in y (mean)
Prevalence
67 146
70 (-45) 10->70 (38.6)
2 (3.0) 3 (2.1)
159 68
20-60 (-52) Adult t
1 (0.6) 2 (2.9)
100 9 58 90
... (54) 17-79 (>50) 0->60 (...) 10-69 (-40)
0 2 (22.2) 1 (1.7) 2 (2.2)
*
(%)
NOTE. All studies involved ambulatory populations. * Ellipses indicate that data are not available. t More specific data are not available.
Several studies, however, report little or no difference between diabetic and nondiabetic groups [19-22]. The conflicting variability in prevalence among reports likely reflects the different populations studied and the varied selection criteria used. As the majority of studies involved patients presenting to an institution, there is also a distinct possibility of selection bias. Diabetics who have more advanced disease with diabetic complications are more likely to require medical care and therefore may be more likely than other diabetics to have bacteriuria. The reported prevalences of asymptomatic bacteriuria among diabetic men have been more consistent (table 2) [12-17, 19-22]; values have ranged from 0.7% to 11.1 % and from 0 to 22.2% for diabetic and nondiabetic men, respectively.Overall, the prevalence is lower among men than among women. Although a few studies suggest an increased prevalence of bacteriuria (symptomatic and asymptomatic) among
diabetic men [12, 16], the majority do not demonstrate a higher prevalence of asymptomatic bacteriuria [13-15, 17, 19-22].
Bacteriology Bacteria isolated from the urine of diabetic and nondiabetic patients with asymptomatic bacteriuria are similar in terms of both the bacterial species involved and the frequency with which each organism is found [12, 14, 19,21,22]. The majority of infections ( rv 75 %) are caused by Escherichia coli. However, other bacteria, including indole-positive and -negative Proteus species, Klebsiella species, Enterobacter species, and Enterococcus faecalis, are frequently cultured. Altogether these organisms are responsible for more than 90 % of cases ofasymptomatic bacteriuria in diabetic and nondiabetic individuals alike.
Downloaded from http://cid.oxfordjournals.org/ at Fachhochschule Osnabrueck Zentrale Einrichtung fuer Wissenschaftliche Infor on April 16, 2015
No. of patients
152
Zhanel, Harding, and Nicolle
Associated Risk Factors
Site of Infection Several investigators have studied the localization of asymptomatic bacteriuria in patients with diabetes mellitus. Ooi et al., using the Fairley bladder washout technique, localized the site of infection to the upper urinary tract in 15 (62.5 %) of 24 diabetic w?men whose ages ranged from 10 to 60 years (mean, 54 years) [15]; the association of age with localization to the upper tract was not discussed. Forland et al. used the antibody-eoated bacteria (ACB) technique to localize the site of infection in diabetic patients [17]. On initial testing, 18 (43 %) of 42 urine specimens were ACB positive, a result suggesting renal parenchymal infection. During 1-16 weeks (mean, 7 weeks) of observation, an additional 15 patients (3~%) converted from ACB negative to ACB positive; thus, ultimately, 33 (79%) of 42 patients were ACBpositive. Forland and Thomas, also using the ACB technique, localized the site of asymptomatic bacteriuria to the upper urinary tract in 32 (71%) of 45 diabetic women [25]. Thus, these three studies suggest that asymptomatic bacteriuria can be localized to the upper urinary tract in more than half of diabetic women.
papillary necrosis in this group [28, 29]. In the report by Ooi et al., all 15 diabetic women with upper urinary tract infection demonstrated impaired urine-eoncentrating ability, whereas the nine patients with lower tract infection retained normal concentrating function [15]. The long-term consequences of this difference (if any) are not clear. Jaspan et al. questioned ?iabetic patients - both with and without bacteriuria- regardmg sy.mptoms suggestive of urinary tract infection (urgency, dysuna, and frequency) [16]. No significant differences were found between the two groups. In a study reported by Batalla et al., diabetics with "persistent [asymptomatic] bacteriuria" or "cured [asymptomatic] bacteriuria" were followed for longer than 40 months to evaluate the contribution of bacteriuria to the development of pyelonephritis and progressive renal damage [27]. No differences were. d~cumented between the two groups in the frequency of clinical symptoms of urinary tract infection, roentgenographic findings of chronic pyelonephritis, abnormal blood urea nitrogen levels, or elevated blood pressures. The investigators concluded that there was no evidence of declining renal function with persistent asymptomatic bacteriuria. Forland and Thomas studied 45 women with diabetes mellitus and asymptomatic bacteriuria [25]. These patients were treated with various protocols based on localization studies and were followed for an average of 34 months. Symptomatic infection developed in four patients, all of whom presented with signs and symptoms suggesting lower urinary tract infection. Seven (16 %) of 45 patients demonstrated significant increases in serum creatinine levels during follow-up; six of the seven patients had proteinuria at the time of enrollment in the study and showed quantitative increases in proteinuria to "3-4+." These investigators concluded that progressive decline of renal function in patients with asymptomatic bacteriuria correlates best with initially present and increasing proteinuria, suggesting that the underlying diabetes per serather than superimposed infection -leads to renal insufficiency. Thus, data from these studies suggest that long-term asymptomatic bacteriuria rarely affects renal function. Whether renal infection can lead to the progression of diabetic renal disease is unclear. In addition, the role of asymptomatic bacteriuria in the diabetic patient in predisposing to sepsis, perinephric abscess, emphysematous cystitis, or papillary necrosis is unknown [10, 28-32].
Morbidity
Treatment
The long-term consequences of asymptomatic bacteriuria in diabetic patients are poorly documented [26]. As previously mentioned, autopsy studies from the preantibiotic era found pyelonephritis to be more common among diabetics than among nondiabetics [2, 3]. In addition, uncontrolled studies reported in 1982 indicated an increased rate of occurrence of renal complications such as perinephric abscess and
Few studies examining the treatment of asymptomatic bacteriuria in patients with diabetes mellitus have been reported [17, 25, 27, 31]. Batalla et al. followed 104 ambulatory diabetics (95 female and nine male) with asymptomatic bacteriuria for a mean of 44 months [27]. The patients' ages ranged from 10 years to >80 years, with almost 75 % of patients over the age of 50. The average duration of documented diabetes
Downloaded from http://cid.oxfordjournals.org/ at Fachhochschule Osnabrueck Zentrale Einrichtung fuer Wissenschaftliche Infor on April 16, 2015
The prevalence of asymptomatic bacteriuria among patients with diabetes mellitus has been correlated with several clinical characteristics. Studies consistently demonstrate that this rate is not influenced by duration or type of diabetes mellitus [3, 12, 16, 23, 27]. It is agreed that the quality of diabetic control (as assessed by glycosylated hemoglobin determinations, mean blood glucose measurements, or mean levels of glycosuria) does not influence the prevalence of asymptomatic bacteriuria [12, 14, 16, 19,22,24]. As diabetic retinopathy becomes more severe, the prevalence of asymptomatic bacteriuria seems to increase [23, 27], although this observation has not been consistent in all studiesIlZ]. The majority of investigations indicate that as renal microangiopathy (assessed by proteinuria and albuminuria) becomes more severe, the prevalence of asymptomatic bacteriuria increases [12, 27]. Again, however, at least one study does not support this observation [14]. No association between decreased renal function and the prevalence of asymptomatic bacteriuria has been described [14, 23, 25]. The prevalence of asymptomatic bacteriuria is reported to be higher among diabetics with heart disease or peripheral vascular disease [23].
RID 1991;13 (January-February)
RID 1991;13 (January-February)
Asymptomatic Bacteriuria in Diabetes
of follow-up. Of these recurrences, reinfectionoccurred more frequently than relapse. Forland and Thomas also studied 45 middle-aged obese Mexican-Americanwomen with non-insulin-dependent diabetes mellitus [25]. The averageage of the patients was 53.5 years, and the average duration of diabetes was 10.8 years. Thirty-one women received insulin, while 14 received oral hypoglycemic agents. Diabetic control was poor in at least half of the study population. Patients had asymptomaticbacteriuria confirmedbyat least two consecutiveculturesof midstream urine that yielded ~108 cfu of the same organism/L. The duration of initial treatment was determined by ACB status. The 13 patients with ACB-negative urine received trimethoprim-sulfamethoxazole (TMP-SMZ, 160/800 mg) twice daily for 2 weeks. Women with ACB-positive results were assigned in alternating fashion to two groups; the first group (18 patients) received TMP-SMZ (160/800 mg) twice daily for 2 weeks, and the second group (14patients)received the same dosage for 6 weeks. Patients with initially positive or negative ACB tests who had ACB-positive urine after the initial 2 weeks of therapy received a 6-week course of therapy. Patientswith ACB-negative reinfections weretreatedwith TMP/SMZ for 2 weeks. Patients in any group with a recurrence after 6 weeks of treatment received low-dosage antimicrobial prophylaxis. Two or moredocumented reinfections served as a second indication for low-dosage antimicrobial prophylaxis. Of the 18 women who received 2 weeks of treatment for ACB-positive infection, 14 (78%) had recurrences, the majority of which were reinfections. Reinfections tended to occur early (within 8 weeks of treatment), while relapses (which occurred in six individuals) tended to occur late (15-127 weeks after therapy). Only four womenhad no recurrences during follow-up (mean, 143 weeks). Among the 14 women who received an initial 6-week course of therapy, 11 recurrences (79%)- nine of them reinfections- and one failure of therapy were documented. Reinfectionsoccurred both early and late, while relapsesoccurred early.Of the 13women who received2 weeksof treatment for initially ACB-negative infection, nine (69%) had recurrences, seven of which were reinfections. Most reinfections occurred late. Eight of the women who received 6 weeks of therapy for ACB-positive recurrence had recurrences within 1 year; six of these recurrences were reinfections. Eleven women received a total of 14 courses of 2-weektreatment for ACB-negative recurrences, with 12 further recurrences (mostly reinfections) after therapy. Reinfections usually occurred late. Nine women received 13courses of low-dosage prophylactic therapy; the duration of prophylaxis was 4-67 weeks. Recurrent infection, usually reinfection, was the rule relatively shortly (average, 7 weeks) after discontinuation of therapy. Four patients experienced seven episodes of symptomatic infection of the lower urinary tract throughoutthe study. Seven(16 %) of 45 patientshad increases
Downloaded from http://cid.oxfordjournals.org/ at Fachhochschule Osnabrueck Zentrale Einrichtung fuer Wissenschaftliche Infor on April 16, 2015
mellitus was 'V15 years; the type of diabetes and whether or not it was insulin dependent were not stated. Baseline renal functionand retinopathywere assessed. After documentation of one positive culture of midstream urine (~108 cfu/L), patients were treated with a variety of antimicrobial agents, includingsulfonamides, chloramphenicol, tetracycline, nitrofurantoin, and methenamine mandelate. (There was no mention of whether methenamine was used with ascorbic acid.) The duration of therapy was not stated. On the basis of the results obtained following therapy, patients with bacteriuria were divided into two groups: those whose infections were cured (34 cases), including patients demonstrating no significantbacteriuria on urine cultures ~1 month after a single course of treatment; and those whose infections persisted (70 cases), including patients with more than one positive culture ~1 month after the completion of initial treatment. Patients were followed for 44 months with approximately one urine culture per year, and assessments of renal function and retinopathy were repeated at the end of the study period. Proteinuria was found with significantly greater frequency at enrollment in the persistent group, and fivepatientsin this group also had significant increasesin proteinuria. There were no such increases in the cured group. Increases in blood urea nitrogen were more common in the persistent group, but this difference was not statistically significant. Retinopathy was documented more frequently in the persistent group at the beginning of the study, and six of seven additional patients reported to have retinopathy at the end of the study were in the persistent group. The authors concluded that microangiopathy usually antedated bacteriuria and may havecontributedto its persistence, although they could not exclude the possibility that bacteriuria playeda role in the progression of microangiopathy. For the duration of follow-up, persistent bacteriuria did not appear to contribute to renal damage. Forland et al. studied 230 womenand 103 men (mean age, 50 years)during 1,127 outpatientreturn clinic visitsin a l-year period [17]. Asymptomaticbacteriuria was diagnosed on the basis of two consecutive cultures showing lOS cfu of the same organismlL. Forty-three women(19%) and two men (2%)had asymptomatic bacteriuria during follow-up. Treatment, chosen on the basis of sensitivity studies, consisted of an oral antimicrobial agentgivenfor 14days. Of the 42 subjectsavailable for continued follow-up, 18 (43%) had initial urine specimens that were ACBpositive. As mentioned earlier, during the pretreatment period (which ranged from 1 to 16 weeks), an additional 15 subjects (36%) converted from ACB negative to ACB positive. Following treatment 22 subjects, 15 (68%) of whom initially had ACB-positive urine, had recurrent infections. Ten(67%)of the 15had reinfections, and five relapsed. Four of the seven ACB-negative individuals with recurrent infections had reinfections, whilethree had relapses. The authors concluded that >50 % of diabetic patients with asymptomatic bacteriuriahad recurrentinfection during 1 year
153
154
Zhanel, Harding, and Nicolle
Summary While asymptomatic bacteriuria is a relatively common finding in diabetic women, the appropriate management of this clinical problem is not clear. Does asymptomatic bacteriuria require antimicrobial treatment in the diabetic, given the frequency of recurrent infection but the lack of data documenting significant morbidity due to bacteriuria? Clinical trials are needed to resolve if and/or when antimicrobial treatment should be given. In addition, long-term studies are needed to characterize the natural history of untreated asymptomatic bacteriuria, particularly with respect to the decline of renal function and the development of symptomatic infection, pyelonephritis, perinephric abscess, papillary necrosis, and septic complications. References 1. Gupta PR, Bajpai HS, Srivastava PK. Urinary tract infection in diabetes mellitus. 1 Indian Med Assoc 1979;72:231-4 2. Sharkey TP, Root HE Infection of the urinary tract in diabetes. JAMA 1935;104:2231-4 3. Baldwin AD, Root HE Infections of upper urinary tract in diabetic patient. N Engl 1 Med 1940;223:224-50 4. MacFarlane lA, Brown RM, Smyth RW, Burdon OW, FitzGerald MG. Bacteraemia in diabetics. J Infect 1986;12:213-9 5. Kass EH. Asymptomatic infections of urinary tract. Trans Assoc Am Physicians 1956;69:56-64 6. Coil lA, Davis JH. Altered host response to experimental pyelonephritis in alloxan diabetic rats. J Surg Res 1967;7:26-34 7. Raffel L, Pitsakis P, Levison SP, Levison ME. Experimental Candida albicans, Staphylococcus aureus, and Streptococcus faecalis pyelonephritis in diabetic rats. Infect Immun 1981;34:773-9 8. Obana Y, Nishino T. The virulence of Enterobacter cloacae and Serratia marcescens in experimental bladder infection in diabetic mice. J Med Microbiol 1989;30:105-9 9. Rayfield EJ, Ault MI, Keusch GT, Brothers M], Nechemias C, Smith
H. Infection and diabetes: the case for glucose control. Am] Med 1982;72:439-50 to. Beyer MM. Diabetic nephropathy. Pediatr Clio North Am 1984;31: 635-51 11. Stamm WE. Measurement of pyuria and its relation to bacteriuria. Am J Med 1983;75(Suppl lB):53-8 12. Hansen ROo Bacteriuria in diabetic and non-diabetic out-patients. Acta Med Scand 1964;176:721-30 13. Vejlsgaard R. Studies on urinary infection in diabetics. I. Bacteriuria in patients with diabetes mellitus and in control subjects. Acta Med Scand 1966;179:173-82 14. JoffeBI, Seftel HC, Distiller LA. Asymptomatic bacteriuria in diabetes mellitus. S Afr Med J 1974;48:1306-8 15. Ooi BS, Chen BTM, Yu M. Prevalence and site of bacteriuria in diabetes mellitus. Postgrad Med J 1974;50:497-9 16. Jaspan lB, Mangera C, Krut LH. Bacteriuria in black diabetics. S Afr Moo ] 1977;51:374-6 17. Forland M, Thomas V, Shelokov A. Urinary tract infections in patients with diabetes mellitus: studies on antibody coating of bacteria. JAMA 1977;238:1924-6 18. SawersIS, ToddWA, Kellett HA, Miles RS, Allan PL, Ewing 0], Clarke BE Bacteriuria and autonomic nerve function in diabetic women. Diabetes Care 1986;9:460-4 19. Schmitt JK, Fawcett C], Gullickson G. Asymptomatic bacteriuria and hemoglobin At. Diabetes Care 1986;9:518-20 20. Huvos A, Rocha H. Frequency of bacteriuria in patients with diabetes mellitus. A controlled study. N Engl J Med 1959;261:1213-6 21. O'Sullivan DJ, FitzGerald MG, MeyneU MJ, Malins JM. Urinary tract infection: a comparative study in the diabetic and general populations. BMJ 1961;1:786-8 22. Abu-Bakare A, Oyaide SM. Asymptomatic bacteriuria in Nigerian diabetics. J Trop Moo Hyg 1986;89:29-32 23. Vejlsgaard R. Studies on urinary infection in diabetics. Il. Significant bacteriuria in relation to long-term diabetic manifestations. Acta Moo Scand 1966;179:183-8 24. Edwards JE Jr, Tillman DB, Miller ME, Pitchon HE. Infection and diabetes mellitus. West] Med 1979;130:515-21 25. Forland M, Thomas VL. The treatment of urinary tract infections in women with diabetes mellitus. Diabetes Care 1985;8:499-506 26. Pometta D, Rees SB, Younger D, Kass EH. Asymptomatic bacteriuria in diabetes mellitus. N Engl J Med 1967;276:1118-21 27. Batalla MA, Balodimos MC, Bradley RF. Bacteriuria in diabetes mellitus. Diabetologia 1971;7:297-301 28. Saiki J, Vaziri NO, Barton e. Perinephric and intranephric abscesses: a review of the literature. West J Med 1982;136:95-102 29. Eknoyan G, Qunibi WY, Grissom RT,]uma SN, Ayus rc Renal papillary necrosis: an update. Medicine (Baltimore) 1982;61:55-73 30. Anderson RU. Urinary tract infections in compromised hosts. Urol Clin North Am 1986;13:727-34 31. Crandon IW, Morgan AG, Williams E. Emphysematouspyelitis and papillary necrosis. West Indian Med J 1982;31:82-5 32. Zhanel GO, HardingGKM, Guay DRP.Asymptomatic bacteriuria: which patients should be treated? Arch Intern Med 1990;150:1389-96 33. Harding GKM, Ronald AR. A controlled study of antimicrobial prophylaxis of recurrent urinary infection in women. N Engl J Med 1974;291:597-601 34. Ronald AR, Harding GKM. Urinary infection prophylaxis in women. Ann Intern Med 1981;94:268-70
Downloaded from http://cid.oxfordjournals.org/ at Fachhochschule Osnabrueck Zentrale Einrichtung fuer Wissenschaftliche Infor on April 16, 2015
in serum creatinine levels to ;?;35 ~mol/L (;?;O.4 mg/dL) during follow-up; all initially had ACB-positive infections. The observations from these studies are compromised by the relatively small numbers of patients. However, the data appear to support the following conclusions: (1) a 2-week course of therapy is equivalent to a 6-week course for the initial eradication of bacteriuria; (2) reinfection occurs more commonly than relapse following treatment of any duration; (3) few patients remain free of bacteriuria for an extended period of follow-up after therapy; and (4) prolonged antimicrobial prophylaxis is effective in preventing bacteriuria, but recurrent infection occurs relatively quickly when prophylaxis is discontinued. This last conclusion is similar to what we have observed in nondiabetic women with recurrent reinfections [33, 34].
RID 1991;13 (January-February)
Asymptomatic Bacteriuria in Patients with Diabetes Mellitus G. G. Zhanel, G. K. M. Harding, and L. E. Nicolle
From the Departments of Medicine and Medical Microbiology, Faculty of Medicine and the Faculty of Pharmacy, University of Manitoba; the Section of Infectious Diseases, St. Boniface General Hospital and Health Sciences Centre; and the St. Boniface General Hospital Research Institute, Winnipeg, Manitoba, Canada
The urinary tract is a frequent site of infection in diabetic patients [1]. Data derived from postmortem evaluationin the preantibiotic era document a histologic prevalenceof "pyelonephritis" of rv 20 % among patients with diabetes mellitus and the occurrence of serious renal infections five times as frequently in diabetic than in nondiabeticpatients [2,3]. Urinary tract infections not only occur more frequently but also are more protracted and have a more serious outcome in diabeticsthan in nondiabetics [3]. Studiesdo notagreeon whether the majority of pyelonephritis cases are the result of hematogenous dissemination or ascending infection, but ascending infectioncertainly contributesa substantialamount to this entity [4, 5]. Studies demonstrate greater susceptibility of diabetic than of nondiabetic animalsto bacteriuriaand pyelonephritis [6-8]. In addition, insulin regulation in diabetic rats does not influence the incidence of infection [6]. Suggested mechanisms for the greater susceptibility of the diabetic urinary tract to
Received for publication30 October 1989and in revised form 24 April 1990. Dr. Zhanel is the recipient of a Postdoctoral Research Fellowship in Medical Microbiology from the Pharmaceutical Manufacturers Association of Canada Health Research Foundation and the Medical Research Council of Canada. The authors thank L. Dingman and S. Singh for their expert secretarial assistance. Please address requests for reprints to Dr. G. G. Zhanel, Department of Microbiology, MS-6, Health Sciences Centre, 820 Sherbrook, Winnipeg, Manitoba, CanadaR3A lR9. Reviews of Infectious Diseases 1991;13:150-4 © 1991 by The University of Chicago. All rights reserved. 0162-0886/91/1301-0067$02.00
infection include decreased antibacterial activity of the urine as a result of dilution of inhibitory substances such as urea, defects in polymorphonuclear leukocyte function or cellular immunity, and increased adhesivecapacity of bladder epithelial cells [6-8]. Several factors may place patients with diabetes mellitus at risk for asymptomaticbacteriuria, including poor control of blood glucoselevels, diabetic neuropathy with neurogenic bladder, impairment of leukocyte function as a result of hyperglycemia, frequent instrumentation of the urinary tract, and diabetic microangiopathy [9, 10]. Asymptomaticbacteriuria is generally definedas the presence of ~108 of a single bacterial species/L (~I()5 cfu/mL) in at least two clean-voided specimens of midstreamurinefrom an individual with no symptoms referable to the genitourinary tract [11]. This review will consider several aspects of asymptomatic bacteriuria in patients with diabetes mellitus, including prevalence, bacteriology, associatedrisk factors, localizationto the upper tract (kidney)or lower tract (bladder), natural history, and treatment.
Prevalence Several studies haveassessed the prevalenceof asymptomatic bacteriuria in women with diabetes mellitus (table I) [12-22]. These studies report a prevalence ranging from 0 to 29.3 % in differentpopulationsof diabetic women. Among nondiabetic control womenthe prevalenceranges from 2.8 % to 22.0 %. The majority of investigatorshavereported a prevalenceofasymptomatic bacteriuriathat is approximately threefoldhigher among womenwith diabetes mellitus than among nondiabetic women [12-26].
Downloaded from http://cid.oxfordjournals.org/ at Fachhochschule Osnabrueck Zentrale Einrichtung fuer Wissenschaftliche Infor on April 16, 2015
This review analyzes several aspects of asymptomatic bacteriuria in patients with diabetes mellitus, including prevalence, bacteriology, coexistent risk factors, localization, natural history, and treatment. The prevalence ofasymptomatic bacteriuria is threefold higher among diabetic women than among nondiabetic women. However, rates among diabetic and nondiabetic men are similar. Studies consistently document that the prevalence of asymptomatic bacteriuria is not influenced by the type or duration of diabetes or by the quality of diabetic control. The microorganisms causing asymptomatic bacteriuria in persons with diabetes mellitus are similar to those causing bacteriuria in nondiabetic individuals. More than half of diabetic patients with asymptomatic bacteriuria have upper urinary tract involvement. The long-term consequences of asymptomatic bacteriuria in patients with diabetes mellitus are poorly documented. Clinical trials describing the treatment of asymptomatic bacteriuria in this population suggest that a 2-week course of therapy is equivalent to a 6-week course for the initial eradication of bacteriuria and that, following treatment, reinfection rather than relapse usually occurs. Important questions remaining include whether asymptomatic bacteriuria should be treated in this population and what the optimal antimicrobial regimen is.
RID 1991;13 (January-February)
Table 1.
Asymptomatic Bacteriuria in Diabetes
151
Prevalence of asymptomatic bacteriuria among women with diabetes mellitus. Diabetic women
Reference
Nondiabetic women
Type of population
Age range in y (mean)
Prevalence
81 128 20 40 152 111 230
70 (-45) 0->70 (31.1) 24-59 (42) 61-82 (68) 10-60 (-55)
341 41 92
Outpatient Outpatient Outpatient Outpatient Outpatient Outpatient Outpatient Outpatient Outpatient Hospitalized Outpatient
15-65 (46) 32-80 (55) 17-79 (>50) 0->60 (...)
15 (18.5) 24 (18.8) 0 8 (20.0) 24 (15.8) 30 (27.0) 43 (18.7) 38 (9.5) 31 (9.1) 12 (29.3) 18 (19.6)
100 41 92
100
Outpatient
10-69 (-40)
9 (9.0)
100
12 13
14 14 15 16 17 18 19 20 21 22
400
Adult" Adult"
(%)
No. of patients
Type of population
81 114 36
Age range in y (mean)
Prevalence
Outpatient Outpatient Outpatient
70 (-45) 10->70 (38) 61-88 (72)
3 (3.7) 9 (7.9) 1 (2.8)
Outpatient Outpatient
20-60 (-55)
(%)
· .. ( ... )
*
152 79
Adult"
7 (4.6) 9 (11.4)
· .. ( ... ) · .. ( ... ) Outpatient Hospitalized Casualty department Outpatient
· .. (54) 17-79 (>50) 0->60 (...)
5 (5.0) 9 (22.0) 17 (18.5)
10-69 (-40)
8 (8.0)
* Ellipses indicate that data are not available. t More specific data are not available.
Table 2.
Prevalence of asymptomatic bacteriuria among men with diabetes mellitus. Diabetic men
Reference
12 13 14 15 16 17 19 20 21 22
Nondiabetic men
No. of patients
Age range in y (mean)
Prevalence
67 141 40 154 87 103 411 9 58 90
70 (45) 0->70 (44.4) 16-77 (54) 20-60 (-52)
5 (7.5) 1 (0.7) 1 (2.5) 2 (1.3) 7 (8.0) 2 (1.9) 4 (1.0) 1(11.1) 2 (3.4) 3 (3.3)
Adultf Adultt
32-80 (55) 17-79 (>50) 0->60 (...) 10-69 (-40)
(%)
No. of patients
Age range in y (mean)
Prevalence
67 146
70 (-45) 10->70 (38.6)
2 (3.0) 3 (2.1)
159 68
20-60 (-52) Adult t
1 (0.6) 2 (2.9)
100 9 58 90
... (54) 17-79 (>50) 0->60 (...) 10-69 (-40)
0 2 (22.2) 1 (1.7) 2 (2.2)
*
(%)
NOTE. All studies involved ambulatory populations. * Ellipses indicate that data are not available. t More specific data are not available.
Several studies, however, report little or no difference between diabetic and nondiabetic groups [19-22]. The conflicting variability in prevalence among reports likely reflects the different populations studied and the varied selection criteria used. As the majority of studies involved patients presenting to an institution, there is also a distinct possibility of selection bias. Diabetics who have more advanced disease with diabetic complications are more likely to require medical care and therefore may be more likely than other diabetics to have bacteriuria. The reported prevalences of asymptomatic bacteriuria among diabetic men have been more consistent (table 2) [12-17, 19-22]; values have ranged from 0.7% to 11.1 % and from 0 to 22.2% for diabetic and nondiabetic men, respectively.Overall, the prevalence is lower among men than among women. Although a few studies suggest an increased prevalence of bacteriuria (symptomatic and asymptomatic) among
diabetic men [12, 16], the majority do not demonstrate a higher prevalence of asymptomatic bacteriuria [13-15, 17, 19-22].
Bacteriology Bacteria isolated from the urine of diabetic and nondiabetic patients with asymptomatic bacteriuria are similar in terms of both the bacterial species involved and the frequency with which each organism is found [12, 14, 19,21,22]. The majority of infections ( rv 75 %) are caused by Escherichia coli. However, other bacteria, including indole-positive and -negative Proteus species, Klebsiella species, Enterobacter species, and Enterococcus faecalis, are frequently cultured. Altogether these organisms are responsible for more than 90 % of cases ofasymptomatic bacteriuria in diabetic and nondiabetic individuals alike.
Downloaded from http://cid.oxfordjournals.org/ at Fachhochschule Osnabrueck Zentrale Einrichtung fuer Wissenschaftliche Infor on April 16, 2015
No. of patients
152
Zhanel, Harding, and Nicolle
Associated Risk Factors
Site of Infection Several investigators have studied the localization of asymptomatic bacteriuria in patients with diabetes mellitus. Ooi et al., using the Fairley bladder washout technique, localized the site of infection to the upper urinary tract in 15 (62.5 %) of 24 diabetic w?men whose ages ranged from 10 to 60 years (mean, 54 years) [15]; the association of age with localization to the upper tract was not discussed. Forland et al. used the antibody-eoated bacteria (ACB) technique to localize the site of infection in diabetic patients [17]. On initial testing, 18 (43 %) of 42 urine specimens were ACB positive, a result suggesting renal parenchymal infection. During 1-16 weeks (mean, 7 weeks) of observation, an additional 15 patients (3~%) converted from ACB negative to ACB positive; thus, ultimately, 33 (79%) of 42 patients were ACBpositive. Forland and Thomas, also using the ACB technique, localized the site of asymptomatic bacteriuria to the upper urinary tract in 32 (71%) of 45 diabetic women [25]. Thus, these three studies suggest that asymptomatic bacteriuria can be localized to the upper urinary tract in more than half of diabetic women.
papillary necrosis in this group [28, 29]. In the report by Ooi et al., all 15 diabetic women with upper urinary tract infection demonstrated impaired urine-eoncentrating ability, whereas the nine patients with lower tract infection retained normal concentrating function [15]. The long-term consequences of this difference (if any) are not clear. Jaspan et al. questioned ?iabetic patients - both with and without bacteriuria- regardmg sy.mptoms suggestive of urinary tract infection (urgency, dysuna, and frequency) [16]. No significant differences were found between the two groups. In a study reported by Batalla et al., diabetics with "persistent [asymptomatic] bacteriuria" or "cured [asymptomatic] bacteriuria" were followed for longer than 40 months to evaluate the contribution of bacteriuria to the development of pyelonephritis and progressive renal damage [27]. No differences were. d~cumented between the two groups in the frequency of clinical symptoms of urinary tract infection, roentgenographic findings of chronic pyelonephritis, abnormal blood urea nitrogen levels, or elevated blood pressures. The investigators concluded that there was no evidence of declining renal function with persistent asymptomatic bacteriuria. Forland and Thomas studied 45 women with diabetes mellitus and asymptomatic bacteriuria [25]. These patients were treated with various protocols based on localization studies and were followed for an average of 34 months. Symptomatic infection developed in four patients, all of whom presented with signs and symptoms suggesting lower urinary tract infection. Seven (16 %) of 45 patients demonstrated significant increases in serum creatinine levels during follow-up; six of the seven patients had proteinuria at the time of enrollment in the study and showed quantitative increases in proteinuria to "3-4+." These investigators concluded that progressive decline of renal function in patients with asymptomatic bacteriuria correlates best with initially present and increasing proteinuria, suggesting that the underlying diabetes per serather than superimposed infection -leads to renal insufficiency. Thus, data from these studies suggest that long-term asymptomatic bacteriuria rarely affects renal function. Whether renal infection can lead to the progression of diabetic renal disease is unclear. In addition, the role of asymptomatic bacteriuria in the diabetic patient in predisposing to sepsis, perinephric abscess, emphysematous cystitis, or papillary necrosis is unknown [10, 28-32].
Morbidity
Treatment
The long-term consequences of asymptomatic bacteriuria in diabetic patients are poorly documented [26]. As previously mentioned, autopsy studies from the preantibiotic era found pyelonephritis to be more common among diabetics than among nondiabetics [2, 3]. In addition, uncontrolled studies reported in 1982 indicated an increased rate of occurrence of renal complications such as perinephric abscess and
Few studies examining the treatment of asymptomatic bacteriuria in patients with diabetes mellitus have been reported [17, 25, 27, 31]. Batalla et al. followed 104 ambulatory diabetics (95 female and nine male) with asymptomatic bacteriuria for a mean of 44 months [27]. The patients' ages ranged from 10 years to >80 years, with almost 75 % of patients over the age of 50. The average duration of documented diabetes
Downloaded from http://cid.oxfordjournals.org/ at Fachhochschule Osnabrueck Zentrale Einrichtung fuer Wissenschaftliche Infor on April 16, 2015
The prevalence of asymptomatic bacteriuria among patients with diabetes mellitus has been correlated with several clinical characteristics. Studies consistently demonstrate that this rate is not influenced by duration or type of diabetes mellitus [3, 12, 16, 23, 27]. It is agreed that the quality of diabetic control (as assessed by glycosylated hemoglobin determinations, mean blood glucose measurements, or mean levels of glycosuria) does not influence the prevalence of asymptomatic bacteriuria [12, 14, 16, 19,22,24]. As diabetic retinopathy becomes more severe, the prevalence of asymptomatic bacteriuria seems to increase [23, 27], although this observation has not been consistent in all studiesIlZ]. The majority of investigations indicate that as renal microangiopathy (assessed by proteinuria and albuminuria) becomes more severe, the prevalence of asymptomatic bacteriuria increases [12, 27]. Again, however, at least one study does not support this observation [14]. No association between decreased renal function and the prevalence of asymptomatic bacteriuria has been described [14, 23, 25]. The prevalence of asymptomatic bacteriuria is reported to be higher among diabetics with heart disease or peripheral vascular disease [23].
RID 1991;13 (January-February)
RID 1991;13 (January-February)
Asymptomatic Bacteriuria in Diabetes
of follow-up. Of these recurrences, reinfectionoccurred more frequently than relapse. Forland and Thomas also studied 45 middle-aged obese Mexican-Americanwomen with non-insulin-dependent diabetes mellitus [25]. The averageage of the patients was 53.5 years, and the average duration of diabetes was 10.8 years. Thirty-one women received insulin, while 14 received oral hypoglycemic agents. Diabetic control was poor in at least half of the study population. Patients had asymptomaticbacteriuria confirmedbyat least two consecutiveculturesof midstream urine that yielded ~108 cfu of the same organism/L. The duration of initial treatment was determined by ACB status. The 13 patients with ACB-negative urine received trimethoprim-sulfamethoxazole (TMP-SMZ, 160/800 mg) twice daily for 2 weeks. Women with ACB-positive results were assigned in alternating fashion to two groups; the first group (18 patients) received TMP-SMZ (160/800 mg) twice daily for 2 weeks, and the second group (14patients)received the same dosage for 6 weeks. Patients with initially positive or negative ACB tests who had ACB-positive urine after the initial 2 weeks of therapy received a 6-week course of therapy. Patientswith ACB-negative reinfections weretreatedwith TMP/SMZ for 2 weeks. Patients in any group with a recurrence after 6 weeks of treatment received low-dosage antimicrobial prophylaxis. Two or moredocumented reinfections served as a second indication for low-dosage antimicrobial prophylaxis. Of the 18 women who received 2 weeks of treatment for ACB-positive infection, 14 (78%) had recurrences, the majority of which were reinfections. Reinfections tended to occur early (within 8 weeks of treatment), while relapses (which occurred in six individuals) tended to occur late (15-127 weeks after therapy). Only four womenhad no recurrences during follow-up (mean, 143 weeks). Among the 14 women who received an initial 6-week course of therapy, 11 recurrences (79%)- nine of them reinfections- and one failure of therapy were documented. Reinfectionsoccurred both early and late, while relapsesoccurred early.Of the 13women who received2 weeksof treatment for initially ACB-negative infection, nine (69%) had recurrences, seven of which were reinfections. Most reinfections occurred late. Eight of the women who received 6 weeks of therapy for ACB-positive recurrence had recurrences within 1 year; six of these recurrences were reinfections. Eleven women received a total of 14 courses of 2-weektreatment for ACB-negative recurrences, with 12 further recurrences (mostly reinfections) after therapy. Reinfections usually occurred late. Nine women received 13courses of low-dosage prophylactic therapy; the duration of prophylaxis was 4-67 weeks. Recurrent infection, usually reinfection, was the rule relatively shortly (average, 7 weeks) after discontinuation of therapy. Four patients experienced seven episodes of symptomatic infection of the lower urinary tract throughoutthe study. Seven(16 %) of 45 patientshad increases
Downloaded from http://cid.oxfordjournals.org/ at Fachhochschule Osnabrueck Zentrale Einrichtung fuer Wissenschaftliche Infor on April 16, 2015
mellitus was 'V15 years; the type of diabetes and whether or not it was insulin dependent were not stated. Baseline renal functionand retinopathywere assessed. After documentation of one positive culture of midstream urine (~108 cfu/L), patients were treated with a variety of antimicrobial agents, includingsulfonamides, chloramphenicol, tetracycline, nitrofurantoin, and methenamine mandelate. (There was no mention of whether methenamine was used with ascorbic acid.) The duration of therapy was not stated. On the basis of the results obtained following therapy, patients with bacteriuria were divided into two groups: those whose infections were cured (34 cases), including patients demonstrating no significantbacteriuria on urine cultures ~1 month after a single course of treatment; and those whose infections persisted (70 cases), including patients with more than one positive culture ~1 month after the completion of initial treatment. Patients were followed for 44 months with approximately one urine culture per year, and assessments of renal function and retinopathy were repeated at the end of the study period. Proteinuria was found with significantly greater frequency at enrollment in the persistent group, and fivepatientsin this group also had significant increasesin proteinuria. There were no such increases in the cured group. Increases in blood urea nitrogen were more common in the persistent group, but this difference was not statistically significant. Retinopathy was documented more frequently in the persistent group at the beginning of the study, and six of seven additional patients reported to have retinopathy at the end of the study were in the persistent group. The authors concluded that microangiopathy usually antedated bacteriuria and may havecontributedto its persistence, although they could not exclude the possibility that bacteriuria playeda role in the progression of microangiopathy. For the duration of follow-up, persistent bacteriuria did not appear to contribute to renal damage. Forland et al. studied 230 womenand 103 men (mean age, 50 years)during 1,127 outpatientreturn clinic visitsin a l-year period [17]. Asymptomaticbacteriuria was diagnosed on the basis of two consecutive cultures showing lOS cfu of the same organismlL. Forty-three women(19%) and two men (2%)had asymptomatic bacteriuria during follow-up. Treatment, chosen on the basis of sensitivity studies, consisted of an oral antimicrobial agentgivenfor 14days. Of the 42 subjectsavailable for continued follow-up, 18 (43%) had initial urine specimens that were ACBpositive. As mentioned earlier, during the pretreatment period (which ranged from 1 to 16 weeks), an additional 15 subjects (36%) converted from ACB negative to ACB positive. Following treatment 22 subjects, 15 (68%) of whom initially had ACB-positive urine, had recurrent infections. Ten(67%)of the 15had reinfections, and five relapsed. Four of the seven ACB-negative individuals with recurrent infections had reinfections, whilethree had relapses. The authors concluded that >50 % of diabetic patients with asymptomatic bacteriuriahad recurrentinfection during 1 year
153
154
Zhanel, Harding, and Nicolle
Summary While asymptomatic bacteriuria is a relatively common finding in diabetic women, the appropriate management of this clinical problem is not clear. Does asymptomatic bacteriuria require antimicrobial treatment in the diabetic, given the frequency of recurrent infection but the lack of data documenting significant morbidity due to bacteriuria? Clinical trials are needed to resolve if and/or when antimicrobial treatment should be given. In addition, long-term studies are needed to characterize the natural history of untreated asymptomatic bacteriuria, particularly with respect to the decline of renal function and the development of symptomatic infection, pyelonephritis, perinephric abscess, papillary necrosis, and septic complications. References 1. Gupta PR, Bajpai HS, Srivastava PK. Urinary tract infection in diabetes mellitus. 1 Indian Med Assoc 1979;72:231-4 2. Sharkey TP, Root HE Infection of the urinary tract in diabetes. JAMA 1935;104:2231-4 3. Baldwin AD, Root HE Infections of upper urinary tract in diabetic patient. N Engl 1 Med 1940;223:224-50 4. MacFarlane lA, Brown RM, Smyth RW, Burdon OW, FitzGerald MG. Bacteraemia in diabetics. J Infect 1986;12:213-9 5. Kass EH. Asymptomatic infections of urinary tract. Trans Assoc Am Physicians 1956;69:56-64 6. Coil lA, Davis JH. Altered host response to experimental pyelonephritis in alloxan diabetic rats. J Surg Res 1967;7:26-34 7. Raffel L, Pitsakis P, Levison SP, Levison ME. Experimental Candida albicans, Staphylococcus aureus, and Streptococcus faecalis pyelonephritis in diabetic rats. Infect Immun 1981;34:773-9 8. Obana Y, Nishino T. The virulence of Enterobacter cloacae and Serratia marcescens in experimental bladder infection in diabetic mice. J Med Microbiol 1989;30:105-9 9. Rayfield EJ, Ault MI, Keusch GT, Brothers M], Nechemias C, Smith
H. Infection and diabetes: the case for glucose control. Am] Med 1982;72:439-50 to. Beyer MM. Diabetic nephropathy. Pediatr Clio North Am 1984;31: 635-51 11. Stamm WE. Measurement of pyuria and its relation to bacteriuria. Am J Med 1983;75(Suppl lB):53-8 12. Hansen ROo Bacteriuria in diabetic and non-diabetic out-patients. Acta Med Scand 1964;176:721-30 13. Vejlsgaard R. Studies on urinary infection in diabetics. I. Bacteriuria in patients with diabetes mellitus and in control subjects. Acta Med Scand 1966;179:173-82 14. JoffeBI, Seftel HC, Distiller LA. Asymptomatic bacteriuria in diabetes mellitus. S Afr Med J 1974;48:1306-8 15. Ooi BS, Chen BTM, Yu M. Prevalence and site of bacteriuria in diabetes mellitus. Postgrad Med J 1974;50:497-9 16. Jaspan lB, Mangera C, Krut LH. Bacteriuria in black diabetics. S Afr Moo ] 1977;51:374-6 17. Forland M, Thomas V, Shelokov A. Urinary tract infections in patients with diabetes mellitus: studies on antibody coating of bacteria. JAMA 1977;238:1924-6 18. SawersIS, ToddWA, Kellett HA, Miles RS, Allan PL, Ewing 0], Clarke BE Bacteriuria and autonomic nerve function in diabetic women. Diabetes Care 1986;9:460-4 19. Schmitt JK, Fawcett C], Gullickson G. Asymptomatic bacteriuria and hemoglobin At. Diabetes Care 1986;9:518-20 20. Huvos A, Rocha H. Frequency of bacteriuria in patients with diabetes mellitus. A controlled study. N Engl J Med 1959;261:1213-6 21. O'Sullivan DJ, FitzGerald MG, MeyneU MJ, Malins JM. Urinary tract infection: a comparative study in the diabetic and general populations. BMJ 1961;1:786-8 22. Abu-Bakare A, Oyaide SM. Asymptomatic bacteriuria in Nigerian diabetics. J Trop Moo Hyg 1986;89:29-32 23. Vejlsgaard R. Studies on urinary infection in diabetics. Il. Significant bacteriuria in relation to long-term diabetic manifestations. Acta Moo Scand 1966;179:183-8 24. Edwards JE Jr, Tillman DB, Miller ME, Pitchon HE. Infection and diabetes mellitus. West] Med 1979;130:515-21 25. Forland M, Thomas VL. The treatment of urinary tract infections in women with diabetes mellitus. Diabetes Care 1985;8:499-506 26. Pometta D, Rees SB, Younger D, Kass EH. Asymptomatic bacteriuria in diabetes mellitus. N Engl J Med 1967;276:1118-21 27. Batalla MA, Balodimos MC, Bradley RF. Bacteriuria in diabetes mellitus. Diabetologia 1971;7:297-301 28. Saiki J, Vaziri NO, Barton e. Perinephric and intranephric abscesses: a review of the literature. West J Med 1982;136:95-102 29. Eknoyan G, Qunibi WY, Grissom RT,]uma SN, Ayus rc Renal papillary necrosis: an update. Medicine (Baltimore) 1982;61:55-73 30. Anderson RU. Urinary tract infections in compromised hosts. Urol Clin North Am 1986;13:727-34 31. Crandon IW, Morgan AG, Williams E. Emphysematouspyelitis and papillary necrosis. West Indian Med J 1982;31:82-5 32. Zhanel GO, HardingGKM, Guay DRP.Asymptomatic bacteriuria: which patients should be treated? Arch Intern Med 1990;150:1389-96 33. Harding GKM, Ronald AR. A controlled study of antimicrobial prophylaxis of recurrent urinary infection in women. N Engl J Med 1974;291:597-601 34. Ronald AR, Harding GKM. Urinary infection prophylaxis in women. Ann Intern Med 1981;94:268-70
Downloaded from http://cid.oxfordjournals.org/ at Fachhochschule Osnabrueck Zentrale Einrichtung fuer Wissenschaftliche Infor on April 16, 2015
in serum creatinine levels to ;?;35 ~mol/L (;?;O.4 mg/dL) during follow-up; all initially had ACB-positive infections. The observations from these studies are compromised by the relatively small numbers of patients. However, the data appear to support the following conclusions: (1) a 2-week course of therapy is equivalent to a 6-week course for the initial eradication of bacteriuria; (2) reinfection occurs more commonly than relapse following treatment of any duration; (3) few patients remain free of bacteriuria for an extended period of follow-up after therapy; and (4) prolonged antimicrobial prophylaxis is effective in preventing bacteriuria, but recurrent infection occurs relatively quickly when prophylaxis is discontinued. This last conclusion is similar to what we have observed in nondiabetic women with recurrent reinfections [33, 34].
RID 1991;13 (January-February)
