Misdiagnosis of Clostridioides Difficile Infections by Standard-of-Care Specimen Collection and Testing Among Hospitalized Adults, Louisville, Kentucky, USA, 2019–20201

Julio A. Ramirez; Frederick J. Angulo; Ruth M. Carrico; Stephen Furmanek; Senén Peña Oliva; Joann M. Zamparo; Elisa Gonzalez; Pingping Zhang; Leslie A. Wolf Parrish; Subathra Marimuthu; Michael W. Pride; Sharon Gray; Cátia S. Matos Ferreira; Forest W. Arnold; Raul E. Istúriz; Nadia Minarovic; Jennifer C. Moïsi; Luis Jodar

Disclosures

Emerging Infectious Diseases. 2023;29(5):919-928.

Abstract and Introduction

Abstract

Although Clostridioides difficile infection (CDI) incidence is high in the United States, standard-of-care (SOC) stool collection and testing practices might result in incidence overestimation or underestimation. We conducted diarrhea surveillance among inpatients ≥50 years of age in Louisville, Kentucky, USA, during October 14, 2019–October 13, 2020; concurrent SOC stool collection and CDI testing occurred independently. A study CDI case was nucleic acid amplification test–/cytotoxicity neutralization assay–positive or nucleic acid amplification test–positive stool in a patient with pseudomembranous colitis. Study incidence was adjusted for hospitalization share and specimen collection rate and, in a sensitivity analysis, for diarrhea cases without study testing. SOC hospitalized CDI incidence was 121/100,000 population/year; study incidence was 154/100,000 population/year and, in sensitivity analysis, 202/100,000 population/year. Of 75 SOC CDI cases, 12 (16.0%) were not study diagnosed; of 109 study CDI cases, 44 (40.4%) were not SOC diagnosed. CDI incidence estimates based on SOC CDI testing are probably underestimated.

Introduction

Clostridioides difficile infection (CDI) is a major cause of illness and death worldwide.^[1,2] The Centers for Disease Control and Prevention (CDC) classifies CDI as an urgent public health threat.^[3] In the CDC Emerging Infections Program (EIP), the CDI incidence in persons ≥50 years of age was 255/100,000 population in 2019, and the hospitalized CDI incidence in this age group was 140/100,000 population.^[4]

CDI incidence estimates derived from public health surveillance rely on standard-of-care (SOC) stool specimen collection and CDI testing practices. Laboratory testing using only a PCR nucleic acid amplification test (NAAT), which tests for the presence of the toxin gene without testing for the presence of free toxin, might misdiagnose a patient with C. difficile carriage as a CDI case-patient and thereby result in overestimation of the CDI incidence.^[5,6] NAAT-alone testing is commonly used by the laboratories in the EIP surveillance sites;^[4] 47% of CDI cases identified in 2017 were diagnosed by a laboratory that used NAAT-alone testing.^[7] Conversely, SOC practices might fail to collect or appropriately test a stool specimen from a person with diarrhea and thereby underdiagnose CDI, which will result in underestimation of CDI incidence.^[8–11]

Incidence estimates are essential for evaluating the need for public health interventions aimed at reducing the CDI burden. We conducted a population-based study to determine CDI incidence and to evaluate the potential effect of misdiagnosis caused by SOC specimen collection and testing practices on CDI incidence estimates in Louisville, Kentucky, USA.

1 2 3 4

Next Section

Abstract and Introduction
Methods
Results
Discussion
References

Table 1. Hospital-based incidence of diarrhea, enrollment rate, stool collection rate, and CDI testing density in participating hospitals before the COVID-19 pause in study of misdiagnosis of CDI by SOC specimen collection and testing among hospitalized adults, Louisville, Kentucky, USA, October 14, 2019–April 11, 2020*
Hospital†	No. adult hospital beds	Surveillance initiation date, 2019	No. eligible patient-days of surveillance	No. incident diarrhea cases identified	Diarrhea incidence‡	No. (%) incident diarrhea cases enrolled	No. (%) study stool specimens collected from enrolled cases	Study CDI stool specimen testing rate, %	Study CDI testing density§
A	671	Oct 14	19,166	195	101.7	163 (83.6)	123 (75.5)	63.1	64.2
B	377	Oct 14	16,252	328	201.8	282 (86.0)	259 (91.8)	79.0	159.4
C	367	Oct 14	11,658	247	211.9	208 (84.2)	183 (88.0)	74.1	157.0
D	348	Oct 14	8,723	143	163.9	122 (85.3)	98 (80.3)	68.5	112.3
E	336	Nov 4	11817	281	237.8	218 (77.6)	133 (61.0)	47.3	112.5
F	200	Dec 23	4,256	41	96.3	32 (78.0)	21 (65.6)	51.2	49.3
G	170	Nov 4	5,754	146	253.7	120 (82.2)	98 (81.7)	67.1	170.3
H	127	Oct 14	8,093	160	197.7	138 (86.3)	132 (95.7)	82.5	163.1
Total	2,596	Oct 14–Dec 23	85,719	1,541	179.8	1,283 (83.3)	1,047 (81.6)	67.9	122.1

*CDI, Clostridioides difficile infection; SOC, standard-of-care. †Participating hospitals are listed in descending order of the number of adult hospital beds.
‡Cases per 10,000 patient-days.
§Tests per 10,000 patient-days.

Table 2. SOC misdiagnosis caused by SOC laboratory testing after excluding recurrent CDI cases in study of misdiagnosis of CDI by SOC specimen collection and testing among hospitalized adults, Louisville, Kentucky, USA, October 14, 2019–April 11, 2020*
SOC CDI test results	No. study primary CDI cases	No. not study primary CDI cases	Total
SOC primary CDI cases	63	12	75
SOC C. difficile carriage	13	17	30
Not SOC primary CDI cases	3	30	33
Total	79	59	138

*CDI, Clostridioides difficile infection; SOC, standard-of-care.

Table 3. Comparison between SOC-diagnosed primary CDI cases (n = 126) and study-diagnosed but not SOC-diagnosed primary CDI cases (n = 44) in study of misdiagnosis of CDI by SOC specimen collection and testing among hospitalized adults, Louisville, Kentucky, USA, October 14, 2019–April 11, 2020*
Characteristic	SOC-diagnosed primary CDI cases, n =126	Study-diagnosed but not SOC-diagnosed primary CDI cases, n = 44†	p value
Demographics
Median age, y, (IQR)	73 (61–81)	67 (59–76)	0.13
Female sex	74 (58.7)	27 (61.4)	0.76
Signs and symptoms
Fever	11 (9.1)	6 (13.6)	0.40
Nausea	38 (34.5)	11 (25.5)	0.29
Abdominal cramping	19 (19.2)	5 (12.8)	0.37
Dehydration	13 (12.5)	2 (5.3)	0.21
Location of diarrhea onset
Hospital-onset‡	22 (17.5)	19 (43.2)	<0.01
Medical history
Dementia	14 (11.1)	4 (9.1)	0.71
Cancer	42 (33.3)	11 (25.0)	0.30
Congestive heart disease	41 (32.5)	23 (52.3)	0.02
Pneumonia	33 (26.2)	13 (29.5)	0.67
Inflammatory bowel disease	17 (13.5)	3 (6.8)	0.24
Diabetes	47 (37.3)	21 (47.7)	0.22
Medical course
Recurrence	4 (3.2)	0 (0)	0.23
PMC	26 (20.8)	0 (0)	<0.01
Transferred to ICU	36 (28.6)	18 (40.9)	0.13
Death at 90 days	20 (15.9)	7 (15.9)	1.00

*Values are no. (%) except as indicated. CDI, Clostridioides difficile infection; ICU, intensive care unit; IQR, interquartile range; PMC, pseudomembranous colitis; SOC, standard-of-care.
†Includes 28 study-diagnosed primary CDI cases that did not have SOC testing and 16 study-diagnosed primary CDI cases that had SOC testing but were not SOC-diagnosed primary CDI cases.
‡Defined as onset of diarrhea ≥3 days after hospital admission.

Table 4. Comparison between enrolled and nonenrolled incident diarrhea cases before the COVID-19 pause (n = 1,490) in study of misdiagnosis of CDI by SOC specimen collection and testing among hospitalized adults, Louisville, Kentucky, USA, October 14, 2019–April 11, 2020*
Characteristic	Enrolled, n =1,249	Nonenrolled, n = 241	p value
Demographics
Median age, y	68	68	NA
Female sex	737 (59)	119 (49)	<0.01
Black	297 (24)	62 (26)	0.52
Medical history
Heart disease	638 (51)	131 (54)	0.35
Cancer	371 (30)	57 (24)	0.06
Stroke	235 (19)	41 (17)	0.51
Dementia	85 (7)	27 (11)	0.02
Antimicrobial drug use in past 2 weeks	370 (30)	74 (31)	0.74
Admission diagnosis
Heart disease	131 (10)	26 (11)	0.89
Respiratory disease	105 (8)	21 (9)	0.88
Pneumonia	95 (8)	20 (8)	0.71
Nervous system disease	61 (5)	21 (9)	0.02
Medical course
Transferred to ICU	297 (24)	73 (30)	0.03
Death in 90 days	152 (12)	53 (22)	<0.01

*Values are no. (%) except as indicated. Surveillance ascertained 1,541 incident diarrhea cases. Medical record review was conducted at 7 of 8 participating hospitals. Hospital F, where 41 incident diarrhea cases were ascertained, did not participate in medical record review. Data from 10 medical records at hospitals A–E, G, and H were not available for this analysis. Data are presented in descending order, where applicable, of Enrolled column. CDI, Clostridioides difficile infection; ICU intensive care unit; NA, not applicable; SOC, standard-of-care.

Table 5. Comparison between inpatients who had a stool specimen collected for CDI testing and inpatients who did not among incident diarrhea cases enrolled before COVID-19 pause (n = 1,283) in study of misdiagnosis of CDI by SOC specimen collection and testing among hospitalized adults, Louisville, Kentucky, USA, October 14, 2019–April 11, 2020*
Characteristic	Study stool specimen collected, n = 1,047	Study stool specimennot collected, n = 236	p value
Demographics
Median age, y (IQR)	68 (61–78)	67.5 (60–77)	0.46
Female sex	595 (56.8)	146 (61.9)	0.16
Black	218 (21.5)	81 (34.8)	<0.01
Hispanic	18 (1.8)	2 (0.9)	0.33
Signs and symptoms
Abdominal pain	283 (28.7)	53 (23.8)	0.14
Fever	109 (10.7)	18 (7.8)	0.18
Blood or pus in stool	90 (9.3)	20 (9.0)	0.91
Medical history
Overnight stay in healthcare facility in previous 12 weeks	615 (60)	150 (63.8)	0.39
Diabetes	432 (41.5)	91 (38.9)	0.46
Antimicrobial drug use in previous 3 months	387 (39.2)	63 (26.9)	< 0.01
Cancer	313 (30.2)	59 (25.1)	0.12
Congestive heart disease	305 (29.3)	64 (27.2)	0.53
CDI in past 5 years	66 (7.5)	18 (8.1)	0.76
Dementia	76 (7.3)	14 (5.9)	0.46
Inflammatory bowel disease	65 (6.3)	10 (4.3)	0.26
Hemiplegia/quadriplegia	54 (5.2)	8 (3.4)	0.25

*Values are no. (%) except as indicated. Data are presented in descending order of Study stool specimen collected column. Values of missing or unknown were excluded from percentage and p value calculations. CDI, Clostridioides difficile infection; IQR, interquartile range; SOC, standard-of-care.

Authors and Disclosures

Julio A. Ramirez², Frederick J. Angulo, Ruth M. Carrico², Stephen Furmanek², Senén Peña Oliva, Joann M. Zamparo, Elisa Gonzalez, Pingping Zhang, Leslie A. Wolf Parrish, Subathra Marimuthu, Michael W. Pride, Sharon Gray, Cátia S. Matos Ferreira³, Forest W. Arnold, Raul E. Istúriz, Nadia Minarovic, Jennifer C. Moïsi and Luis Jodar

University of Louisville, Louisville, Kentucky, USA (J.A. Ramirez, R.M. Carrico, S. Furmanek, S. Peña Oliva, L.A. Wolf Parrish, S. Marimuthu, F.W. Arnold); Pfizer Vaccines, Collegeville, Pennsylvania, USA (F.J. Angulo, J.M. Zamparo, E. Gonzalez, P. Zhang, S. Gray, C.S. Matos Ferreira, R.E. Istúriz, N. Minarovic, J.C. Moïsi, L. Jodar); Pfizer, Pearl River, New York, USA (M.W. Pride)

¹Study results were presented at the annual Anaerobe 2022 Congress, July 28–31, 2022, Seattle, Washington, USA.

²Current affiliation: Norton Healthcare, Louisville, Kentucky, USA.

³Current affiliation: AstraZeneca Pharmaceuticals, Wilmington, Delaware, USA.

Address for correspondence
Frederick J. Angulo, Medical Development and Scientific/Clinical Affairs, Pfizer Vaccines, 500 Arcola Rd, Collegeville, PA 19426, USA; email: frederick.j.angulo@pfizer.com

About the Author
Dr. Ramirez is chief scientific officer for the Norton Infectious Diseases Institute at Norton Healthcare in Louisville, KY, and emeritus professor of medicine, Division of Infectious Diseases, University of Louisville. His primary research interests include respiratory tract infections, with a focus on community-acquired pneumonia and emerging respiratory pathogens.

This study was supported by Pfizer Inc., which was involved in the study design, data collection, data analysis, data interpretation, and writing of the article. All authors had full access to the study data and approved the decision to submit for publication. Upon request, and subject to review, Pfizer will provide the data that support the findings of this study. Subject to certain criteria, conditions and exceptions, Pfizer may also provide access to the related individual de-identified participant data. For more information, see https://www.pfizer.com/science/clinical-trials/trial-data-and-results.
J.A.R. has a grant from Pfizer Inc. to the University ofLouisville Research Foundation for the epidemiologic study.F.J.A., J.M.Z. E.G., P.Z., M.W.P., S.G., R.E.I., J.C.M, and L.J.hold stock and stock options in Pfizer Inc. R.M.C. has grantsor contracts from Pfizer Inc. for C. difficile research andhonoraria from Pfizer Inc. for content development andpresentation regarding C. difficile to professionalorganization. L.A.W.P. and S.M. declare support from PfizerInc. for this study (i.e., kits for performing Quik Chek rapidtests, supplies for aliquoting stool samples, sample forms, and barcoding labels and scanners); contract from Pfizer Inc.to support the project; and receipt of equipment (i.e.,1 biologic safety cabinet, 1 refrigerator, 1 incubator, 1anaerobic chamber, and 1 freezer). C.S.M.F. declaressupport from Pfizer Inc. for this manuscript and stockoptions from AstraZeneca, PLC. F.W.A. declaresinstitutional payment from Pfizer Inc. for this manuscriptand grants or contracts from the US Health Resources andServices Administration for the Ryan White HIV clinic, Janssen for COVID-19 vaccine study, and Gilead for rapidstart antiretroviral therapy study. The other co-authorsdeclare no conflicts of interest.