SOAP note AHA Week 7

UncomplicatedUTIinwomen.NP.12.pdf

Home >Nursing homework help >SOAP note AHA Week 7

www.tnpj.com The Nurse Practitioner • May 2012 41

ymptoms of urinary tract infection (UTI) remain a common presenting problem among women in primary care. Approximately 9 million women per

year are seen in offi ces, clinics, or EDs for UTIs, at a cost of over $2.47 billion.1 The lifetime prevalence of at least one occurrence of UTI is estimated at over 50% for women.1 Diagnosis and treatment based on a woman’s demographic characteristics and symptoms, sometimes combined with urine dipstick tests for nitrites and leukocyte esterase (LE), are the mainstay of clinical management and consistent with contemporary guidelines by the American College of Obste- tricians and Gynecologists.2 However, the relative inaccuracy of clinical diagnosis of UTIs continues to be raised in the literature, especially due to heightened concerns about unnec- essary treatment with antibiotics and antibiotic-resistant bacteria.1,3-5

The authors conducted a study of 263 military women with either urinary symptoms or both urinary and vaginal symptoms, and systematically evaluated the accuracy of urinary symptoms, urine dipstick tests, and combinations of the two tests in comparison to the standard diagnosis of UTI via urine culture. The study also evaluated advanced practice registered nurses’ (APRN) clinical diagnoses of UTI compared to urine culture results.

Despite current practice recommendations that the diagnosis of lower UTI is based primarily on signs, symp- toms, and patient history, the gold standard for diagnosis of UTI is a clean-catch urine culture positive for uropatho- gens.6 Given the cost and time delay for results of urine cultures, a range of empirical diagnostic methods is

employed including dipstick urinalysis.6 Research shows that the amount and type of information used to diagnose women with urinary symptoms has no signifi cant impact on diagnostic accuracy.

By Nancy K. Lowe, PhD, CNM, FACNM, FAAN, and Nancy A. Ryan-Wenger, PhD, RN, CPNP, FAAN

Uncomplicated UTIs in women

Abstract: Empirical diagnosis and treatment of lower urinary tract infection (UTI)

in women is the most common clinical approach due to the urgency of symptoms and cost.

This study examines the importance of recognizing common symptoms and accurately

diagnosing UTIs in the primary care setting.

Key words: diagnosis of urinary tract infection, urinary tract infection, urine dipstick tests, urine culture.

Ph ot

o ©

zi lli

/i st

oc kp

ho to

.c om

Uncomplicated UTIs in women

42 The Nurse Practitioner • Vol. 37, No. 5 www.tnpj.com

Empirical diagnosis and treatment of urinary symptoms The effectiveness of telephone management of urinary symptoms is rarely evaluated by urine culture because of the diffi cult logistics of obtaining urine specimens. Safety and effectiveness are typically evaluated by the percentage of women who experience continued or recurrent symp- toms after empirical treatment. For example, within 8 weeks (N = 273) and 6 weeks (N = 4,177) after empirical treatment by telephone triage, the percentage of women who made subsequent offi ce visits for recurrent or persistent symptoms was 16.8% (n = 46) and 17.9% (n = 748), respectively.7,8 Pyelonephritis was diagnosed in 2.2% (n = 6), and 0.5% (n = 21), respectively. In a third study, 72 women with uri- nary symptoms who called for a primary care provider appointment were randomized to telephone management (n = 36) or offi ce management (n = 36).9

All of the women were diagnosed and treated for UTI, and all but fi ve provided a urine specimen for analysis and culture prior to taking antibiotics. Positive urine cultures were found in 64.2% (n = 43) of the women. The two groups were not signifi cantly different on symptom scores, satisfac- tion with care, or urine culture results. The authors noted that the percentage of positive cultures in women diagnosed with UTI was similar to that of a study published in 1980, but did not express concern that nearly one-third of the women had negative cultures, indicating that they were misdiagnosed and treated with antibiotics unnecessarily. Even when women are diagnosed and treated in offi ce or clinic settings, the accuracy of providers’ diagnoses is not unfailing. For example, in a sample of 111 women with urinary symptoms, 68 (61.3%) received antibiotics, but only 27 (39.7%) had positive urine cultures, and of the 43 (38.7%) women who were not treated, 9 (20.9%) had pos- itive urine cultures.10 The scope of information used by these providers to make their diagnoses was not described.

Diagnostic value of urinary symptoms and point-of-care tests The search for better diagnostic tools for UTI prompted research on the accuracy of urinary symptoms and point- of-care tests in the diagnosis of UTI compared to the gold standard of urine culture. Results from nine research articles were summarized in a systematic review.11 The highest positive likelihood ratios (LR+) of single observations were achieved from dipstick urinalysis (LR+ = 4.2) and self- diagnosis based on previous symptoms of UTI (LR+ = 4.0), compared to symptoms of hematuria, urinary frequency, costovertebral angle pain, back pain, fever, or dysuria (LR+ = 2.0, 1.8, 1.7, 1.6, and 1.6, respectively).

Many studies show similar diagnostic accuracy levels for urine nitrites (N) and leukocyte esterase (LE) dipstick

tests compared to urine culture. The results of two studies showed that in a sample of 1,583 women with urinary symptoms, compared to urine culture of 100 colony-forming units (CFUs)/mL or greater, the presence of N had a posi- tive predictive value (PPV) of 90%, and a negative predic- tive power (NPV) of 30%. When N was negative, the PPV of LE+ was 79%.4 This indicates that theoretically, a positive LE test will “catch” 79% of UTI cases when nitrites are negative. In another study of 408 women with symptoms of UTI, the odds ratios for signs and symptoms were high- est for N+ at 6.36, and leukocytes (4.52), followed by dysuria (2.76), blood (2.23), protein (1.12), and urgency (1.1). Combinations of signs and symptoms did not improve the odds ratio for N+, although no data were provided to sup- port this statement.12

Comparison of accuracy rates across studies is compli- cated by the use of different statistics (LR+, odds ratio, PPV, NPV, sensitivity, specifi city) and lack of a standard defi nition for a positive urine culture. For example, the above studies defi ne UTI as greater than 102 CFU/mL greater than 103 CFU/mL, or greater than 105 CFU/mL of a single pathogenic organism. As the CFU standard increases, the number of UTI diagnoses decrease, the number of non-UTI diagnoses increase, and the risk for false negative diagnoses increase. In addition, sample inclusion and exclusion criteria varied among studies. Women with vaginal symptoms accompany- ing urinary symptoms were often excluded. Despite the fact that in clinical practice women often present with both urinary and vaginal symptoms, the authors found no stud- ies within the last 20 years that reported the accuracy of UTI diagnoses for women with and without vaginal symptoms. The authors’ current study extends knowledge about the diagnostic value of urinary symptoms and urine dipstick tests among women who seek care for urinary symptoms alone or in combination with vaginal symptoms. Previous research has evaluated providers’ accuracy in the diagnosis of UTI under various circumstances such as telephone man- agement and in office settings where the knowledge of patient history, physical exam, and point-of-care testing used to make clinical diagnoses is varied.

This study was unique in that experienced APRNs fol- lowed a systematic protocol that included all of these recom- mended sources of data and that the authors compared the accuracy of their clinical diagnoses to urine culture in women with and without vaginal symptoms. The research questions were: 1. What is the accuracy (PPV and NPV) of urinary symp-

toms, dipstick urinalysis (N and LE), and their combina- tions, compared to urine culture in the diagnosis of UTI in symptomatic women with or without vaginal symp- toms?

Uncomplicated UTIs in women

www.tnpj.com The Nurse Practitioner • May 2012 43

2. What is the accuracy of APRNs’ clinical diagnoses of UTI compared to urine culture in symptomatic women with or without vaginal symptoms?

■ Methods Setting and sample This report is a secondary analysis of data from the authors’ study on the accuracy of self-diagnosis of genitourinary infection in 715 military women.13 The participants included in this analysis are the subset of military women (N = 263) who presented at a United States Army or Navy troop med- ical clinic with one or more of the classic symptoms of UTI (burning, urgency, or frequency). The participants ranged from 18 to 54 years of age (mean = 26.4, median = 24, mode = 21), and nearly half were married/living with a signifi cant other (n = 124, 47.4%) and one-third were single (n = 87, 33.3%) (see Demographic characteristics).

Consistent with the demographics of military women, the racial and ethnic distribution was 41% White, 30% Black, and 20.1% Hispanic, and rank distribution was 1/4 enlisted, 1/3 noncommissioned offi cer, and less than 10% commis- sioned offi cers.

Procedure The procedure for the parent study has been previously described in detail.13 A relevant aspect of this report is that women presenting with at least one symptom of UTI who consented to participate were seen by one of four masters- prepared APRNs who were nationally certifi ed in women’s health, and had 6 to 11 years of experience in that role. The APRNs used a standardized clinical research protocol to con- duct a clinical interview and evaluate clinical signs through a physical exam and assessment of a clean-catch urine specimen for N and LE. The APRNs made their own clinical diagnoses at the time of exam and treated the women as appropriate. Clean-catch urine specimens from all the women were sent to the hospital lab for culture. A positive urine culture was defi ned as 105 CFU/mL or greater of uropathic organisms, consistent with traditional criteria and past research.2,10,11 Follow-up for untreated positive urine culture results occurred as needed.

Analysis The accuracy of urinary symptoms, urine N and LE, and their combinations in the diagnosis of culture-confi rmed UTI was measured in terms of PPV and NPV, and their 95% confi dence intervals. Ideally, both PPV and NPV should approach 100%. These statistics refl ect the realities of clin- ical practice, such as the probability that symptoms and signs will accurately differentiate UTIs and non-UTIs confi rmed by urine culture. The accuracy of APRN’s clinical diagnoses was evaluated further for error in treatment decisions.

Results Of the 263 participants, 153 (58.2%) presented with urinary symptoms alone and 110 (41.8%) presented with a combina- tion of urinary and vaginal symptoms. Overall, there were 137 positive urine cultures representing 10 different uropathic microorganisms, of which Escherichia coli (E. coli) was the most common (n = 103, 75.2%) (see Frequency and percent- age of uropathic microorganisms cultured at ≥105 CFU/mL).

Predictive value of urinary symptoms and dipstick urinalysis One hundred forty-three women presented with only uri- nary symptoms. Among 50 women with negative N and LE, there were 14 positive urine cultures (28%), and for the 7 women who had positive N only, all 7 had a positive urine culture. Sixty-four women had a positive LE only, of whom

Demographic characteristics (N = 263)

Demographic Characteristic n %

Race/Ethnicity

American Indian/Alaska Native or 1 0.4

Native Hawaiian/Pacifi c Islander 2 0.8

Asian 10 3.8

Black 75 28.7

White 114 43.7

Hispanic 51 19.5

More than one race 8 3.1

Education

High School Graduate or GED 50 19.1

Some College 151 57.6

Associate Degree 28 10.7

College Graduate 28 10.7

Post-Graduate 5 1.9

Marital Status

Single 87 33.2

Married/Living with signifi cant other 124 47.4

Divorced/Separated 50 19.1

Military Branch

Army 124 47.1

Navy 127 48.3

Air Force/Marine/Coast Guard 12 4.6

Military Rank

Enlisted 67 25.6

Noncommissioned Offi cer 174 66.4

Warrant Offi cer/Junior Offi cer 17 6.5

Senior Offi cer 3 1.1

Uncomplicated UTIs in women

44 The Nurse Practitioner • Vol. 37, No. 5 www.tnpj.com

49 (76.6%) had positive urine cultures, while 32 women had positive N and LE, of whom 27 (84.4%) had positive cultures (see Women presenting with urinary symptoms only).

One hundred ten women presented with both urinary and vaginal symptoms. Among the 49 women with a negative urine dipstick for both N and LE, there were 8 (16.3%) positive urine cultures. Only 3 women had positive N only, and 2 of the urine cultures were positive. Among the 41 women with LE positive only, there were 15 (36.6%) positive urine cultures, and for the 17 women with both N and LE positive on dipstick, there were 14 (82.4%) positive urine cultures (see Women presenting with urinary and vaginal symptoms).

The data were further analyzed to determine the PV and NPV of urine culture for symptoms alone, urine dipstick results alone, and urine dipstick in combination with symp- toms (see Accuracy of symptoms, urine dipstick, and combina- tions to predict urine culture). Only combinations with a frequency of 10 or more cases were analyzed. In the entire sample of 263 participants, one-fourth (n = 66, 25.1%) of the women experienced only one of the three classic urinary symptoms, 29.7% (n = 78) had two symptoms, and 44.9% (n=118) had the symptom triad of burning, urgency, and frequency. The PPV of a single symptom or combination of other symptoms ranged from 28% for only frequency symp- toms to a high of 66.1% for the symptom triad. NPVs are important because they indicate the probability that the absence of selected symptoms and/or negative dipstick results would be consistent with a negative culture result. NPV was fairly stable at 45.4% to 48.7%, with the exception of the symptom triad (59.4%). The 95% confi dence intervals

were wide for all of the symptoms, ranging from 17.7% to 58.9% for PPVs and 12.4% to 15.9% for NPVs. A narrow confi dence interval minimizes the potential margin of diag- nostic error (see Positive and negative predictive values).14

While it is unlikely that urine dipstick would be the only information used to diagnose UTIs, it is instructive to note that with positive LE alone and in combination with positive N, the PPV increases to 84% and 100%, respectively, but these fi ndings occurred in only about one-fourth of the women in the study. NPVs were similar to those found with symptoms alone, and again, 95% confi dence intervals were wide for both PPVs and NPVs. When symptoms and urine dipstick results were combined, only an estimated 44% of the sample had one of the four combinations. The most accurate prediction of positive culture occurred when all three classic symptoms were present, and the urine was positive for both N and LE (PPV = 89.1%). However, only 29 (11.5%) of the entire sample had this combination. NPVs were similar to those for symptoms and urine dipstick results alone, and 95% confi dence intervals were wide.

■ Accuracy of APRNs’ clinical diagnosis of UTI Of the 263 diagnoses made by APRNs, 65.7% (n = 173) were confi rmed by urine culture. APRNs accurately diagnosed 124 (47.1%) women with UTI and 49 (18.6%) women with non-UTI. The false positive rate was 29.3%, indicating that 77 women were unnecessarily treated with antibiotics (com- mission errors of treatment), and the false negative rate was 4.9%, indicating that 13 women did not receive appropriate antibiotics for their UTI infection (omission errors of treat- ment) at the time of their clinical evaluation. Diagnostic accuracy rates were similar between the two groups of women without vaginal symptoms (64.7%, n = 90) and with vaginal symptoms (67.3%, n = 74). False positive and false negative rates for the two groups were also similar to those of the entire group. It is important to note that the women with both urinary and vaginal symptoms were less likely to have a culture-confi rmed UTI (30.9%) as opposed to women with urinary symptoms only (58.8%).

■ Discussion The fi ndings support current knowledge about the relatively poor diagnostic value of urinary symptoms alone, urine N and LE alone, or in combination. The study protocol required APRNs to conduct a standardized clinical history, physical exam, and clean-catch dipstick urinalysis before making a diagnosis using their own clinical judgment. The research setting also guaranteed that the APRNs had suffi cient time to conduct the clinical protocol in an uninterrupted manner without time constraints. Despite the additional time and clinical information, only approximately two-thirds of their

Frequency and percentage of uropathic

microorganisms cultured at ≥ 105 CFU/mL.

Microorganism Frequency Percent

No growth 116 45.8

E. coli 103 40.7

Group B streptococcus 7 2.7

Staphylococcus saprophyticus 6 2.4

Proteus mirabilis 6 2.4

Citrobacter koseri 5 2.0

Klebsiella pneumoniae 5 2.0

Enterobacter aerogenes 2 0.8

Citrobacter freundii 1 0.4

Gamma Streptococcus group D enterococcus

1 0.4

Serratia species 1 0.4

Total 253 100

Uncomplicated UTIs in women

www.tnpj.com The Nurse Practitioner • May 2012 45

Women presenting with urinary symptoms only

The fl ow chart represents the urine dipstick results, urine culture results, and clinical diagnoses of UTI for women with urinary symptoms only.

N = nitrites, LE = leukocyte esterase, CFU = colony-forming units/mL, APRN = Advanced Practice RN, UTI = urinary tract infection.

Accurate diagnosis False positive diagnosis False negative diagnosis of UTI

APRN clinical diagnosis of UTI prior to urine culture results

No n = 36

(23.5%)

Yes n = 14 (9.2%)

No n = 0

Yes n = 7

(4.6%)

No n = 15 (9.8%)

Yes n = 49 (32%)

No n = 5

(3.3%)

Yes n = 27

(17.6%)

Yes n = 30 (19.6%)

No n = 6

(3.9%)

Yes n = 10 (6.5%)

No n = 4

(2.6%)

Yes n = 7

(4.6%)

No n = 0

Yes n = 13 (8.5%)

No n = 2

(1.3%)

Yes n = 47

(30.7%)

No n = 2

(1.3%)

Yes n = 4

(2.6%)

No n = 1

(0.6%)

Yes n = 26 (17%)

No n = 1

(0.6%)

No n = 7

(4.6%)

LE+

n = 64 (41.8%)

N+LE+ n = 32

(20.9%)

N–LE–

n = 50 (32.7%)

N = 153

Women presenting with urinary symptoms only

Clean-catch urine dipstick results

Uropathic culture results

CFU 105 CFU 105 CFU 105 CFU 105

Uncomplicated UTIs in women

46 The Nurse Practitioner • Vol. 37, No. 5 www.tnpj.com

Women presenting with urinary and vaginal symptoms

The fl owchart represents the urine dipstick results, urine culture results, and clinical diagnoses of UTI for women with both urinary and vaginal symptoms.

N = nitrites, LE = leukocyte esterase, CFU = colony-forming units/mL, APRN = Advanced Practice RN, UTI = urinary tract infection.

Accurate diagnosis False positive diagnosis False negative diagnosis

APRN clinical diagnosis of UTI prior to urine culture results

No n = 41

(37.2%)

Yes n = 8

(7.3%)

No n = 0

Yes n = 3

(2.7%)

No n = 26

(26.3%)

Yes n = 15 (1.5%)

N+

n = 3 (2.7%)

Yes n = 14

(12.7%)

Yes n = 10 (9%)

No n = 31

(28.2%)

Yes n = 5

(4.5%)

No n = 3

(2.7%)

Yes n = 2

(1.8%)

No n = 1

(0.9%)

Yes n = 17

(15.5%)

No n = 9

(8.2%)

Yes n = 14

(12.7%)

No n = 1

(0.9%)

Yes n = 3

(2.7%)

No n = 0

Yes n = 13

(11.8%)

No n = 1

(0.9%)

N+

N = 3 (2.7%)

LE+

N = 41 (37.3%)

N+L+ N = 17 (15.5%)

N–LE–

N = 49 (44.5%)

N = 110

Women presenting with urinary and vaginal symptoms

Clean-catch urine dipstick results

Uropathic culture results

CFU ≥ 105 CFU ≥ 105 CFU ≥ 105 CFU ≥ 105

Uncomplicated UTIs in women

www.tnpj.com The Nurse Practitioner • May 2012 47

diagnoses were accurate when evaluated against the gold standard of urine culture. An estimated 30% of the women were erroneously diagnosed and treated for UTI with anti- biotics, a fi nding similar to the results of other studies.

Regardless of the amount and type of information used to diagnose UTI, error rates range from 25.8% to 39.7%.9,10 This is a concern because the unnecessary use of antibiotics contributes to the development of antibiotic-resistant organisms. Only about 5% of the APRN’s diagnoses of non- UTI were in error, meaning that 5% of the women actually had a UTI but were not treated for it. The authors have extended the science of UTI diagnosis by demonstrating that a comprehensive clinical history and physical exam does not appreciably improve diagnostic accuracy of UTI in women with urinary symptoms with and without vaginal symptoms. The presence of vaginal symptoms does not interfere with the accuracy of UTI or non-UTI diagnoses, but it is important to note that the accurate diagnosis of non-UTI was six times greater in women with vaginal symp- toms (36.4%) than women with only urinary symptoms only (5.9%).

This study confirms that despite the actions of well- trained and experienced clinicians with adequate time for a full clinical evaluation, the accurate diagnosis of UTI from clinical symptoms and signs remains diffi cult. APRNs were more likely to over treat for UTI (29.3% commission errors of treatment) than under treat (4.9% omission errors of treat- ment) when using a diagnostic urine culture criterion of 105

CFU/mL. Therefore, with a goal to reduce unnecessary treat- ment with antibiotics, data suggest that in both women pre- senting with symptoms of UTI alone or in combination with vaginal symptoms and negative urine dipsticks for both N and LE, an appropriate clinical consideration is to postpone antibiotic treatment until culture results are known since only 28% and 16.3% of these women had positive cultures. Depending on the population and the individual woman’s sexual history, screening for sexually transmitted infections (STIs) may also be appropriate. However in this sample, in which STI screening was done on all participants per protocol, only 7 women (2.8%) had chlamydia and 2 had gonorrhea (less than 1%). Urethritis rather than cystitis (UTI) may be the explanation for the presence of clinical symptoms and may be amenable to symptomatic treatment unless it is related to chlamydia. Until culture results are known, symptomatic treatment with pyridium, cranberry extracts, and increased fl uids can be initiated.

The study data did not include the investigation of more subtle aspects of clinical assessment and diagnostic reasoning. Data were not collected to allow the evaluation of severity of symptoms, other related signs and symptoms, or rapid onset that the APRNs may have used in their clinical decision mak- ing. The APRNs undoubtedly incorporated additional clin- ical data such as internal versus external dysuria, onset of dysuria, small voids, gross hematuria, and suprapubic pain or tenderness into their clinical decisions. However, treat- ment errors as determined by urine culture still occurred.

Accuracy of symptoms, urine dipstick, and combinations to predict urine culture

Symptoms N (%) PPV (CI) NPV (CI)

Burning only 31 (11.4%) 29.0% (14.9–48.2) 45.5% (39.1–52.0)

Urgency only 10 (3.7%) 60.0% (27.4–86.3) 48.7% (42.5–54.9)

Frequency only 25 (9.2%) 28.0% (12.9–49.6) 46.0% (39.7–52.4)

Burning and urgency 17 (6.2%) 52.9% (28.5–76.1) 48.4% (42.2–54.7)

Burning and frequency 12 (4.4%) 58.3% (28.6–83.5) 48.7% (42.5–54.9)

Urgency and frequency 49 (17.9%) 42.9% (29.1–57.7) 46.4% (39.8–53.2)

Burning, urgency, and frequency 118 (43.2%) 66.1% (56.7–74.4) 59.4% (51.2–67.1)

Urine dipstick

LE only 107 (39.2%) 60.7% (50.8–69.9) 54.2% (46.3–61.9)

N only 10 (3.7%) 100% (65.5–100) 50.2% (44.0–56.4)

LE and N 50 (18.3%) 84.0% (70.3–92.4) 55.6% (48.8–62.2)

Negative LE and negative N 106 (38.8%) 22.6% (15.3–32.0) 29.9% (23.2–37.6)

Combination of symptoms with positive N and/or LE

Burning, urgency, and LE 10 (3.7%) 70.0% (35.4–91.9) 49.0% (42.9–55.2)

Urgency, frequency, and LE 21 (7.7%) 52.4% (30.3–73.6) 48.4% (42.1–54.8)

Burning, urgency, frequency, and LE 53 (19.4%) 69.8% (55.5–81.3) 52.7% (45.9–59.4)

Burning, urgency, frequency, N and LE 29 (10.6%) 89.7% (71.5–97.3) 52.9% (46.4–59.2)

Uncomplicated UTIs in women

48 The Nurse Practitioner • Vol. 37, No. 5 www.tnpj.com

Consistent with the body of the clinical literature, E. coli was the most predominant organism associated with UTI. Therefore, APRNs must continue to teach women and young girls the essentials of protective perineal hygiene and other health behaviors known to reduce the incidence and/or reoccurrence of UTIs.

The findings also support the need for a systematic detailed history when evaluating women either over the phone or in person with symptoms of UTI. While empiric treatment is common and considered a cost-effective standard of care with or without dipstick urinalysis, further research is indicated to identify whether a systematic analysis of sever- ity of symptoms can improve diagnostic accuracy particularly with a goal to reduce unnecessary antibiotic treatment.

REFERENCES 1. Griebling TL. Urologic diseases in America project: trends in resource use for

urinary tract infections in women. J Urol. 2005;173(4):1281-1287.

2. American College of Obstetricians and Gynecologists (ACOG). Treatment of urinary tract infections in nonpregnant women. Obstet Gynecol. 2008; 111(3):785-794.

3. Little P, Turner S, Rumsby K, et al. Dipsticks and diagnostic algorithms in urinary tract infection: development and validation, randomised trial, eco- nomic analysis, observational cohort and qualitative study. Health Technol Assess. 2009;13(19):iii-iv, ix-xi, 1-73.

4. Nys S, van Merode T, Bartelds AI, Stobberingh EE. Urinary tract infections in general practice patients: diagnostic tests versus bacteriological culture. J An- timicrob Chemother. 2006;57(5):955-958. Epub 2006 Mar 22.

5. Ellington MJ, Livermore DM, Pitt TL, Hall LM, Woodford N. Mutators among CTX-M beta-lactamase-producing Escherichia coli and risk for the

emergence of fosfomycin resistance. J Antimicrob Chemother. 2006;58(4):848- 852. Epub 2006 Aug 5.

6. National Guideline Clearinghouse. Guideline Synthesis: Diagnosis and Man- agement of Uncomplicated Lower Urinary Tract Infection. Rockville, MD: Na- tional Guideline Clearinghouse; 2009. http://www.guideline.gov.

7. Schauberger CW, Merkitch KW, Prell AM. Acute cystitis in women: experi- ence with a telephone-based algorithm. WMJ. 2007;106(6):326-329.

8. Vinson DR, Quesenberry CP. The safety of telephone management of pre- sumed cystitis in women. Arch Intern Med. 2004;164:1026-1029.

9. Barry HC, Hickner J, Ebell MH, Ettinhofer T. A randomized controlled trial of telephone management of suspected urinary tract infections in women. J Fam Pract. 2001;50(7):589-594.

10. O’Brien K, Hillier S, Simpson S, Hood K, Butler C. An observational study of empirical antibiotics for adult women with uncomplicated UTI in general practice. J Antimicrob Chemother. 2007;59(6):1200-1203. Epub 2007 Apr 21.

11. Bent S, Nallamothu BK, Simel DL, Fihn SD, Saint S. Does this woman have an acute uncomplicated urinary tract infection? JAMA. 2002;287(20):2701-2710.

12. Little P, Turner S, Rumsby K, et al. Developing clinical rules to predict urinary tract infection in primary care settings: sensitivity and specifi city of near pa- tient tests (dipsticks) and clinical scores. Br J Gen Pract. 2006;56(529):606- 612.

13. Ryan-Wenger NA, Neal JL, Jones AS, Lowe NK. Accuracy of vaginal symptom self-diagnosis algorithms for deployed military women. Nurs Res. 2010;59(1): 2-10.

14. Straus SE, Richardson WS, Glasziou P, Haynes RB. Evidence-based Medicine: How to Practice and Teach EBM. 3rd ed. Edinburgh: Elsevier, Churchill, Living- stone, 2005.

This study was funded by the National Institutes of Health, National Institute of Nursing Research.

Nancy K. Lowe is a professor and division chair of Women, Children, and Family Health at the University of Colorado College of Nursing in Aurora, Colo.

Nancy A. Ryan-Wenger is the director of nursing research at Nationwide Children’s Hospital in Columbus, Ohio.

The authors have disclosed that they have no fi nancial relationships related to this article.

DOI-10.1097/01.NPR.0000413483.52003.f8

Positive and negative predictive values

P e

rc e

n ta

g e

0 n=25 n=31 n=49 n=17 n=12 n=10 n=118 n=21 n=29n=10n=53

Fr eq

uen cy

o nly

Burn in

g o nly

Urg en

cy &

fr eq

uen cy

Burn in

g & u

rg en

Burn in

g & fr

eq uen

Urg en

cy o

nly

Urg en

cy , f

re quen

ct &

leu ko

cy te

es te

ra se

Sx T ria

d & le

uko cy

es te

ra se

Burn in

g, u rg

en cy

leu ko

cy te

es te

ra se

Burn in

g u rg

en cy

fre quen

ct (S

x T ria

Sx T ria

d, le uko

cy te

es te

ra se

+ &

N itr

ite s +

100

Positive Predictive Value

Negative Predictive Value

Sx = Symptom