Against Screening Urinalyses for Asymptomatic Bacteriuria in Children

The Case

Kathi J. \s=b\

Kemper, MD, MPH, Ellis

Screening children for asymptomatic bacteriuria to prepyelonephritis and renal scarring is widely recom-

vent

mended, but its cost-effectiveness has not been established.

published studies to determine the costs and benefits of screening toilet-trained, asymptomatic children for bacteriuria. Given the sensitivity and specificity of current screening methods (approximately 80% each) and the prevalence of bacteriuria in asymptomatic children (approximately 1% in girls and 0.03% in boys), screening 100 000 children would result in 19897 (20%) false\x=req-\ positive tests; initial screening and two urine cultures to confirm the diagnosis of asymptomatic bacteriuria would miss 28% of 515 children with true bacteriuria, and cost nearly $2.9 million. There is no evidence that detection and treatment of children with asymptomatic bacteriuria prevents subsequent pyelonephritis or renal scarring. Screening for bacteriuria in asymptomatic children is costly, fails to prevent pyelonephritis or renal scarring, and should be discontinued as a part of routine well-child care. We reviewed

(AJDC. 1992;146:343-346)

are widely performed on asymp¬ as a part of routine health supervi¬ tomatic sion. Although periodic screening is recommended by the American Academy of Pediatrics1-2 and by authors of sev¬ eral general pediatrie textbooks,3-4 others do not recom¬ mend that such screening be performed on asymptomatic outpatients5"11 or during routine hospital admissions.12"14 Screening urinalyses generally include examinations for the presence of glucose, protein, blood, and bacteria. Because of short lag time, eg, between glucosuria and the onset of symptoms of diabetes and high false-positive rates eg, proteinuria and hematuria, there is general con¬ sensus that screening for glucosuria, proteinuria, and he¬ maturia in asymptomatic children is not costeffective.7-8-15"19 However, screening is still promoted on the premise that early detection and treatment of asymp¬ tomatic bacteriuria (ABU) can prevent pyelonephritis, re-

children Screening urinalyses

Accepted

publication October 7, 1991. From the Departments of Pediatrics, University of Washington (Drs Kemper and Avner), Children's Hospital and Medical Center for

(Dr Avner), and Harborview Medical Center (Dr Kemper), Seattle, Wash. Reprint requests to Harborview Medical Center, Department of Pediatrics, Room ZA-53, 325 Ninth Ave, Seattle, WA 98104 (Dr Kem-

per).

D.

Avner, MD

nal

scarring, and end-stage renal disease. In practice, screening for ABU generally has been restricted to chil¬ dren who have been toilet-trained due to problems with contamination in specimens collected in bags and the cost and discomfort in obtaining specimens with a catheter or with suprapubic aspiration. LITERATURE REVIEW of the controversy in our institutions about whether Because to continue to screen for ABU, we reviewed published studies20 to answer several questions: (1) what is the accuracy of current screening tests in detecting ABU?; (2) what is the cost of screen¬ ing tests in terms of false-positive and false-negative results and dollars charged?; and (3) does early detection of ABU reduce the risk of pyelonephritis and renal scarring? How Accurate Are Current Tests in Detecting ABU? Screening tests are generally assessed with respect to their sensitivity (the ability to detect true positive results), specificity (the ability to detect true negative results), positive predictive value (the likelihood that a person with a positive screening test result truly has the condition), and negative predictive value (the likelihood that a person with a negative screening test result does not have the condition) compared with a diagnostic "gold stan¬ dard." In the case of ABU, screening test results are compared with findings on consecutive urine cultures. The accuracy of a screening test varies according to the defi¬ nition of the gold standard, such as the number of colonies of pure culture (slOOOO vs 5=100000) and the number of consecu¬ tive cultures that are required to meet the criteria (one vs two vs three). The more stringent the gold standard (higher colony counts and more consecutive cultures required), the higher the false-positive rate of the screening test. To avoid missing cases of potentially serious conditions, the cutoff for a positive result

screening test is generally set to maximize sensitivity, usu¬ ally at the cost of specificity, again increasing the number of false positives. on a

Different types of tests have been used to

screen

for ABU,

including dipstick and microscopic examination. Several authors2123 have reported that some dipstick tests are as accurate and less expensive in screening for ABU than microscopic

examination of urinary sediment. In a recent review of screen¬ ing for ABU in adults, Pels et al24 noted that leukocyte esterase has acceptable sensitivity and specificity for a screening test in adults (Table 1). Similarly, Goldsmith and Campos,25 in screen¬ ing children for ABU, found that leukocyte esterase is as accu¬ rate as sediment microscopy compared with urine culture. Even using the most rigorous methods for specimen collection and analysis, the highest estimates for sensitivity and specificity for

Downloaded From: http://archpedi.jamanetwork.com/ by a University of Calgary User on 05/25/2015

Table 2.—Variation in Predictive Values With Differing Prevalence of ABU in Boys and Girls With Constant Sensitivity and Specificity of 80%*

1.—Sensitivity and Specificity of Screening Tests for Bacteriuria

Table

Test

Leukocyte

esterase

Microscopic examination for

white blood cells Microscopic examination for bacteria •Review of studies in adult standards.

Sensitivity, %

Specificity, %

72-87*

64-82*

76t

81t

82t

81t

80t

83t

Present

growth.

urine screening using leukocyte esterase or microscopic exami¬ nation are about 80%. The predictive value of a screening test is affected by the prevalence of the condition in the population as well as the test's sensitivity and specificity. The more common the condition, the higher the positive predictive value will be. The prevalence of ABU varies with the gender, age, and race of the screened population. Among boys, the prevalence varies from a high of 2.5% in infancy (before true screening is currently feasible) to 0% to 0.14% at school-age.26"29 Studies among girls indicate that the highest rate of ABU is found among sexually active adolescents, eg, 3.5% in black teenagers,28 and the lowest rate is found in infancy and preschool, 0.8% to 0.9%.26-27-30'31 Studies reporting higher rates in school-age girls, eg, 1.2% to

1.9%, frequently included girls who were symptomatic.28-31,32 Overall, the best estimate of the prevalence of ABU among toilet-trained, preschool boys is approximately 0.03% and

girls is approximately 1%. a sensitivity and specificity of 80% with dipstick or mi¬ croscopic screening and a true prevalence of ABU of 0.03% in boys and 1% in girls, the positive predictive value of an initial screen is 0.1% in boys and 3.9% in girls (Table 2). The negative predictive values are greater than 99% in both sexes. among

Given

What Is the Cost of Screening? The cost of screening tests can be measured for false-positive results and false-negative results as well as the financial charges for performing the initial tests and subsequent diagnostic eval¬ uations. It is difficult to assign a dollar amount to the psycholog¬ ical and emotional burdens imposed by falsely labeling a child as having a disease or being at increased risk of a disease, but these effects have been well described in numerous studies33"36 of conditions associated with the development of the vulnerable child syndrome. False-negative test results, on the other hand, can lead to false reassurance that no further evaluation is needed and delay diagnosis while the disease progresses. Given the prevalence of ABU in preschool children and the sensitivity and specificity of current screening tests, falsepositive results would occur in approximately 9997 (20%) of 50000 boys and 9900 (20%) of 50000 girls during the initial screening test (Table 2). Thus, nearly 20000 children of every 100 000 screened would be falsely labeled as having ABU until further diagnostic tests were performed. Positive screening test results are generally confirmed with at least two urine cultures.37 If the first culture is negative, further evaluations are not performed. However, if the first culture is positive, it is repeated to minimize the number of false-positive test results, unnecessary antimicrobial therapy, and additional diagnostic evaluations. One urine culture has a sensitivity and specificity of approximately 95% for detecting ABU on subse¬ quent cultures.38 The number of false-negative results in the hypothetical cohort of 100000 children can be calculated given the sensitivity of the tests and the prevalence of ABU in the population. With a sen-

Boys (True Prevalence of ABU 0.03%)* 9997 12 Positive screening 39988 3 Negative test

10009

Total

50000

15

49 985

Girls (True Prevalence of ABU = 1%)t 400 9900 Positive screening

patients using several different gold

with S10000 colonies per milliliter of pure

Total

=

tStudy in pediatrie patients using gold standard of one urine cul¬

ture

Absent

in

Negative test

100

Total

500

39600 49500

39991

10300

39700 50 000

*ABU indicates asymptomatic bacteriuria. tThe positive predictive value is 0.1% in boys (12/10009) and 3.9%

girls (400/10300).

Table 3.—Costs of Initial Screening Test and Urine Cultures in Hypothetical Cohort of 100000 Children Cost per Test

screening test boys and 50 000 gi rls)

Total Cost

Initial

(50000

First urine culture (10009 boys and 10300 girls)

Second urine culture (511 boys and 875

girls)

Total

$6.25-$19.50

$625 000-$1 950 000

$21.25-$24

$431566-$487416

$21.25-$24

$29425-$33264 $1 085 991 -$2 470 680

sitivity of 80%, the initial screening test would have missed three boys and 100 girls with ABU (Table 2). With a sensitivity of 95%,

the first culture would miss an additional 21 children and the second culture would miss an additional 19 children for a total of 143 children with true ABU who would be missed with this screening strategy. This represents an overall false-negative rate of 28% of 515 children with true ABU in this population. Although the cost of urine dipsticks is minimal, clinic, labora¬ tory, and hospital charges for performing urinary screening can be substantial (Table 3). At the Children's Hospital and Medical Center, Seattle, Wash, in 1990, the cost of a routine urinalysis (which automatically included microscopic examination) was $19.50; at the Harborview Medical Center Children's Clinic, Se¬ attle, Wash, the charge for a urine dipstick screening test was $6.25 and the charge for microscopic examination of the urine was $15. Total charges for initial screening for 100000 children could vary from $625 000 to $1950 000, depending on the site and method of screening. Given the results of the initial screening and local costs of $21.25 at Harborview Medical Center to $24 at Children's Hos¬ pital and Medical Center for each urine culture, initial diagnos¬ tic charges for our hypothetical cohort of 100 000 children can be calculated as in Table 3. The overall charges range from $1085 991 to $2470680 at our institutions. These figures represent initial evaluations only and do not include the cost of physician visits, therapeutic or prophylactic antimicrobial therapy, diagnostic imaging, or days missed from school or work.

Does

Early Detection and Treatment of ABU Reduce the Risk of Pyelonephritis and/or Renal Scarring?

Numerous studies have addressed the natural history of ABU in children with and without underlying renal disease, eg, scar-

Downloaded From: http://archpedi.jamanetwork.com/ by a University of Calgary User on 05/25/2015

other parenchymal abnormalities, and reflux. In girls with ABU or "covert" bacteriuria without underlying renal abnormalities who are not treated with antimicrobial ther¬ apy, bacterial isolates tend to remain stable for several years.39 Although bacteriuria tends to persist if left untreated, fewer than 10% of patients develop symptomatic urinary tract infections and fewer than 5% of patients develop pyelonephritis.40'41 How¬ ever, if patients are given antimicrobial therapy, bacterial isolates change in more than 70% of patients, symptomatic urinary tract infection develops in approximately 10% of patients, and pyelo¬ nephritis develops in 5% to 18% of patients.28-39-4 4 The results of three randomized controlled trials of antimicrobial treatment in patients with ABU or covert bacteriuria demonstrated that treatment did not improve the prognosis for subsequent symp¬ tomatic infections, reflux, or renal growth, and perhaps in¬ creased the risk of developing pyelonephritis.45-17 Therefore, the value of screening asymptomatic normal children with the aim of treating bacteriuria is highly uncertain. Estimates of the underlying prevalence of renal abnormalities vary from 1.37% in infants48 to 0.5% in schoolchildren with problems detected on screening ultrasound examinations.49 Among infants and children with ABU, approximately 9% to 11% have reflux and 13% have upper urinary tract disease at the time ABU is detected.26-27 The proportion of such abnormalities in children with ABU is approximately half of that detected in children with symptomatic urinary tract infections.26-50"53 Al¬ though renal scarring can occur in older children, most scars oc¬ cur in the first 2 to 3 years of life, before children are toilet-trained and screening is feasible.26-54"56 Therefore, detection of ABU to identify children with underlying renal abnormalities and pre¬ vent renal scarring is inefficient and poorly justified. In girls with ABU and renal scarring or vesicoureteral reflux who were not treated with antimicrobial therapy, only 6% developed further scarring during 4 years of follow-up.47 Al¬ though treatment may temporarily eradicate ABU, most girls have recurrences within 1 year28; continuous treatment is asso¬ ciated with breakthrough pyelonephritis in from 0.4 per 100 to 1.8 per 100 patient-months of therapy.57-58 In the study by Hansson et al43 of asymptomatic girls with renal scarring, the risk of pyelonephritis was similar in girls whether they were treated or not (two of 12 vs 0 of 14). Such studies indicate that antimicro¬ bial therapy of even high-risk children with ABU may be

ring,

unwarranted.

COMMENT The realization that so many abnormalities associated with systemic disease could be easily identified through urinalysis led our clinical predecessors to recommend that this examination be widely performed to screen for underlying disease. Later assessments of the risks and benefits of screening for proteinuria, glucosuria, and hematuria suggested that such screening should not be per¬ formed in asymptomatic children.12 However, the hope that serious problems, such as renal scarring and endstage renal disease, could be prevented through early identification and treatment of asymptomatic infections has led to the persistence of recommendations for peri¬ odic screening of bacteriuria in asymptomatic children as a part of routine well-child care. However, the present analysis demonstrates that be¬ current screening techniques are only moderately accurate and ABU is relatively uncommon, especially in boys, current screening recommendations lead to a high rate of false-positive test results, many false-negative re¬ sults, and a high financial cost for initial evaluations. Fur¬ thermore, most renal scarring occurs before screening urinalyses are really feasible with current techniques. There is no evidence known to us that treatment of ABU cause

is beneficial in

preventing pyelonephritis or further renal

scarring, even in children with underlying renal abnor¬ malities. In fact, because antimicrobial treatment can substantially alter the bacterial flora associated with ABU, such treatment may increase the risk of subsequent infec¬ tion and scarring. Thus, we believe that periodic screen¬ ing for asymptomatic bacteriuria in children should not be recommended as a part of routine well-child care. Future efforts should be directed toward developing costeffective screening techniques to prevent and treat pyelo¬ nephritis during infancy and early childhood. References 1. Committee on Standards of Child Health Care. Standards of Child Health Care. 3rd ed. Evanston, Ill: American Academy of Pediatrics; 1977:9-36. 2. Guidelines for Health Supervision II. Evanston, Ill: American

Academy of Pediatrics;

1988:155-159.

3. Mortimer EA. Preventive

pediatrics

and

epidemiology.

In: Behr-

RE, Vaughan VC, eds. Nelson's Textbook of Pediatrics. 13th ed. Philadelphia, Pa: WB Saunders Co; 1987:158. 4. Wiebe RA. Ambulatory care. In: Kelley VC, ed. Practices of Pediatrics. Philadelphia, Pa: Harper & Row Publishers Inc; 1987;1:1-16. 5. Lawrence RS, Mickalide AD. Preventive services in clinical practice: designing the periodic health examination. JAMA. 1987;257:2205\x=req-\ man

2207. 6. Spitzer WO. The periodic health examination, I: introduction. Can Med Assoc J. 1984;130:1276-1285. 7. Silver HK, Kempe CH, Bruyn HB, Fulginiti VA, eds. Handbook of Pediatrics. 15th ed. East Norwalk, Conn: Appleton & Lange; 1987:441. 8. Hoekelman RA, ed. Primary Pediatric Care. St Louis, Mo: Mosby\p=n-\ Year Book; 1987:29. 9. Zhanel GG, Harding GK, Guay DR. Asymptomatic bacteriuria: which patients should be treated? Arch Intern Med. 1990;150:1389-1395. 10. Pless IB. Routine tests in pediatric practice: things better left un-

done. Pediatr 11.

Dig. 1979;21:13. Dodge WF. Cost-effectiveness of renal disease screening. AJDC.

1977;131:1274-1280. 12. Akin BV, Hubbell FA, Frye EB, Rucker L, Friis R. Efficacy of routine admission urinalysis. Am JMed. 1987;82:719-722. 13. Lawrence VA, Kroenke K. The unproven utility of pre-operative urinalysis. Arch Intern Med. 1988;148:1370-1373.

14. Mitchell N, Stapleton FB. Routine admission urinalysis examination in pediatric patients: a poor value. Pediatrics. 1990;86:345-348. 15. Dodge WF, West EF, Smith EH, Bunce H. Proteinuria and hematuria in school children: epidemiology and early natural history. JPediatr. 1976;88:327-347. 16. Gutgesell M. Practicality of screening urinalyses in asymptomatic children in a primary care setting. Pediatrics. 1978;62:103-105. 17. Vehaskari VM, Rapola J, Kaskimies O, Savilahti E, Vilska J, Hallman N. Microscopic hematuria in schoolchildren: epidemiology and clini-

copathologic evaluation. J Pediatr. 1979;95:676-684. 18. Vehaskari VM, Rapola J. Isolated proteinuria: analysis of a school-age population. J Pediatr. 1982;101:661-668. 19. Woolhandler S, Pels RJ, Bor DH, Himmelstein DU, Lawrence RS. Dipstick urinalysis screening of asymptomatic adults for urinary tract disorders, I: hematuria and proteinuria. JAMA. 1989;262:1215-1219. 20. Newman TB, Browner WS, Halley SB. The case against childhood cholesterol screening. JAMA. 1990;264:3039-3043. 21. Mariani AJ, Luangphinith S, Loo S, Scottolini A, Hodges CV. Dipstick chemical urinalysis: an accurate cost-effective screening test. J Urol. 1984;132:64-66. 22. Kusumi RK, Penelope JG, Kunin CM. Rapid detection of pyuria by leukocyte esterase activity. JAMA. 1981;245:1653-1655. 23. Hamoudi AC, Bubis SC, Thompson C. Can the cost savings eliminating urine microscopy in biochemically negative urines be extended to

the

Am JClin Pathol. 1986;86:658-660. RJ, Bor DH, Woolhandler S, Himmelstein DU, Lawrence RS.

pediatric population?

24. Pels

Dipstick urinalysis screening of asymptomatic adults for urinary tract disorders. JAMA. 1989;262:1221-1224. 25. Goldsmith BM, Campos JM. Comparison of urine dipstick, microscopy, and culture for the detection of bacteriuria in children. Clin Pediatr. 1990;29:214-218. 26. Jodal U. The natural history of bacteriuria in childhood. Infect Dis Clin North Am. 1987;1:713-729. 27. Siegel SR, Siegel B, Sokoloff BZ, Kanter MH. Urinary infection in infants and preschool children. AJDC. 1980;134:369-372. 28. Kunin CM, Deutscher R, Paquin A. Urinary tract infection in school children: an epidemiological, clinical and laboratory study. Medicine. 1964;43:91-130. 29. Silverberg DS. City-wide screening for urinary abnormalities in school boys. Can Med Assoc J. 1974;111:410-412. 30. Freeman JW, Sindhu SS. A survey for bacteriuria in schoolgirls.

Downloaded From: http://archpedi.jamanetwork.com/ by a University of Calgary User on 05/25/2015

Med J Aust1974;1:135-137. . 31. Savage DCL. Natural history of covert bacteriuria in school girls. Kidney Int. 1975;8(suppl):90-95. 32. Silverberg DS, Allard MJ, Ulan RA, et al. City-wide screening for urinary abnormalities in schoolgirls. Can Med Assoc J. 1973;109:981-985. 33. Kemper K, Forsyth B, McCarthy P. Jaundice, terminating breast\x=req-\ feeding, and the vulnerable child. Pediatrics. 1989; 84:773-778. 34. Bodegaard G, Fyro K, Larsson A. Psychological reactions in 102 families with a newborn who has a falsely positive screening test for congenital hypothyroidism. Acta PaediatrScand. 1983;304(suppl):15-20. 35. Rothenberg MB, Sills EM. latrogenesis: the PKU anxiety syndrome. J Am Acad Child Psychiatry. 1986;7:689-692. 36. Bergman AB, Stamm SJ. The morbidity of cardiac nondisease in schoolchildren. N Engl J Med. 1967;276:1008-1013. 37. Rapkin RH. Urinary tract infection in childhood. Pediatrics.

1977;60:508-511. 38. Kass EH, Finland M. Asymptomatic infections of the urinary tract. Trans Assoc Am Physicians. 1956;69:56-64. 39. Hansson S, Caugant D, Jodal U, Svanborg-Eden C. Untreated asymptomatic bacteriuria in girls, I: stability of urinary isolates. BMJ. 1989;298:853-855. 40. Lindberg U, Claesson I, Hansson LA, Jodal U. Asymptomatic bacteriuria in school girls, VII: clinical course during a 3-year follow-up. J Pediatr. 1978;92:194-199. 41. Newcastle Covert Bacteriuria Research Group. Covert bacteriuria in school girls in Newcastle-Upon-Tyne: a 5-year follow-up. Arch Dis

Child. 1981;56:585-592. 42. Gillenwater JY, Harrison RB, Kunin CM. Natural history of bacteriuria in school girls. N Engl J Med. 1979;301:396-399. 43. Hansson S, Jodal U, Noren L, Bjure J. Untreated bacteriuria in as1989;84:964-968. ymptomatic girls with renal scarring. Pediatrics. 44. Kunin CM. The natural history of recurrent bacteriuria in school Med. N 1970;282:1443-1448. girls. Engl J 45. Savage DCL, Howie G, Adler K, Wilson Ml. Controlled trial of bacteriuria in childhood. Lancet. 1975;1:358-361. in therapy covert

46. Lindberg U. Asymptomatic bacteriuria in school girls, V: the clinical course and response to treatment. Acta Paediatr Scand.

1975;64:718-724.

47. Cardiff-Oxford Bacteriuria Study Group. Sequelae of covert bacteriuria in school girls. Lancet. 1978;1:889-893. 48. Steinhart JM, Kuhn JP, Eisenberg B, Vaughan RL, Maggioli AJ, Cozza TF. Ultrasound screening of healthy infants for urinary tract ab-

normalities. Pediatrics. 1988;82:609-614. 49. Sheih CP, Liu MB, Hung CS, Yang KH, Chen WY, Lin CY. Renal abnormalities in school children. Pediatrics. 1989; 84:1086-1090. 50. Stork JE. Urinary tract infection in children. Adv Pediatr Infect Dis.

1987;2:115-134. 51. Shah KJ, Robins DG, White RH. Renal scarring and vesicoureteral reflux. Arch Dis Child. 1978;53:210-217. 52. Tappin DM, Murphy AV, Mocan H, et al. A prospective study of children with first acute symptomatic E coli urinary tract infection: early 99 m technetium dimercaptosuccinic acid scan appearances. Acta Paediatr Scand. 1989;78:923-929. 53. Strife JL, Bisset GS, Kirks DR, et al. Nuclear cystography and renal sonography: findings in girls with urinary tract infection. AJR Am J Roentgenol. 1989:153:115-119. 54. Smellie JM, Ransleg PG, Normand ICS, Prescod N, Edwards D. Development of new renal scars: a collaborative study. BMJ.

1985;290:1957-1960. 55. Winter AL, Hardy BE, Alton DJ, Arbus GS, Churchill BM. Acquired renal scars in children. J Urol. 1983;129:1190-1194. 56. Randolph MF, Morris KE, Gould EB. The first urinary tract infection in a female infant. J Pediatr. 1975;86:342-348. 57. Hanson E, Hansson S, Jodal U. Trimethoprim-sulfadiazine prophylaxis in children with vesico-ureteric reflux. Scand J Infect Dis. 1989;21:201-204. 58. Jodal U, Larsson P, Hansson S, Bauer CA. Pivmecillinam in long\x=req-\ term prophylaxis to girls with recurrent urinary tract infection. Scand J Infect Dis. 1989;21:299-302.

In Other AMA

Journals

ARCHIVES OF SURGERY A National Survey of Surgeons' Attitudes About Patients With Human Immunodeficiency Virus Infections and

Acquired Immunodeficiency Syndrome Gene Ann

Shelley, PhD, Richard J. Howard, MD, (Arch Surg. 1992;127:206-212)

PhD

Double Gloving: Protecting Surgeons From Blood Contamination in the Operating Room Edward J. Quebbeman, MD, PhD; Gordon L. Telford, MD; Karen Wadsworth, RN; Susan Hubbard, RN, BSN, CNOR; Hannah Goodman, MS; Mark S. Gottlieb, PhD

(Arch Surg. 1992;127:213-217)

The Surgical Infection Society's Policy on Human Immunodeficiency Virus and Hepatitis and Hepatitis C Infection: The Ad Hoc Committee on Acquired Immunodeficiency Syndrome and Hepatitis John Mihran Davis, MD; Robert H. Demling, MD; Frank R. Lewis, MD; Eddie Hoover, MD; J. Paul Waymack, MD

(Arch Surg. 1992;127:218-221)

Downloaded From: http://archpedi.jamanetwork.com/ by a University of Calgary User on 05/25/2015