Congenital Cytomegalovirus Testing Outcomes from the ValEAR Trial

Abstract

Objective:

To determine the positivity rate of congenital cytomegalovirus (cCMV) testing among universal, hearing targeted CMV testing (HT-cCMV) and delayed targeted dried blood spot testing (DBS) newborn screening programs, and to examine the characteristics of successful HT-cCMV testing programs.

Study Design:

Prospective survey of birth hospitals performing early CMV testing.

Methods:

Birth hospitals participating in the NIH funded ValEAR clinical trial were surveyed to determine the rates of cCMV positivity associated with three different testing approaches: universal testing, HT-cCMV, and DBS testing. A mixed methods model was created to determine associations between successful HT-cCMV screening and specific screening protocols.

Results:

Eighty-two birth hospitals were surveyed from February 2019 to December 2021. Seven thousand six hundred seventy infants underwent universal screening, 9017 infants HT-cCMV and 535 infants delayed DBS testing. The rates of cCMV positivity were 0.5%, 1.5% and 7.3% respectively. The positivity rate for universal CMV screening was less during the COVID-19 pandemic than that reported prior to the pandemic. There were no statistically significant drops in positivity for any approach during the pandemic. For HT-cCMV testing, unique order sets and rigorous post-testing protocols were associated with successful screening programs.

Conclusion:

Rates of cCMV positivity differed among the three approaches. The rates are comparable to cohort studies reported in the literature. Universal CMV prevalence decreased during the pandemic but not significantly. Institutions with specific order set for CMV testing where the primary care physician orders the test and the nurse facilitates the testing process exhibited higher rates of HT-cCMV testing.

Article summary:

Results of a multi-institutional study evaluating testing approaches for congenital cytomegalovirus infection. Early diagnosis is vital for the implementation of time sensitive management for pediatric hearing loss.

INTRODUCTION

Congenital cytomegalovirus (cCMV) is the most common non-genetic cause of SNHL in the United States (US).¹ The prevalence of cCMV infection in the United States, Canada, Australia and Western Europe is estimated to be between 0.5% and 0.7% of all live births.^2–4 Infants with cCMV are at an increased risk for sensorineural hearing loss (SNHL), and long-term neurologic sequelae. ^5–7 However, cCMV is rarely diagnosed at birth in the absence of a screening program owing to the often clinically invisible nature of the infection. ^5–7 While 10–15% of neonates with cCMV are born with visible signs, the majority (85–90%) are born without signs or symptoms making the diagnosis exceedingly challenging. ^{8 9–11} Early diagnosis of cCMV is important to address any sequelae present at birth, and to monitor for future sequelae, including SNHL, which occurs in up to 25% of infants with asymptomatic infections.¹² Adding to the challenge of diagnosing cCMV is the need to test using a specimen collected prior to 3 weeks of age, after which point it can be very difficult to differentiate a congenital infection from a postnatal infection, which is not associated with neurologic sequelae.^13,14

Recently, CMV testing due to legislative mandates and a recognition to diagnose this condition have become increasingly more common, resulting in more frequent diagnosis and early interventions for affected infants. Debate remains about the ideal method of neonatal testing for cCMV. Universal newborn screening programs, in which all infants are screened at birth, identify all cases of cCMV, yet some argue that such programs would be challenging and costly to implement, and cause undue anxiety to families since the vast majority will not develop clinically significant disease.¹⁵ To date most universal cCMV screening programs have utilized dried blood spot (DBS) testing, which has a lower sensitivity when compared to other specimens such as saliva and urine for PCR testing (~70% versus 99%).¹⁶ An alternative approach consists of testing all neonates who fail their newborn hearing screen for cCMV, known as hearing targeted CMV screening (HT-cCMV), which identifies a subset of those with cCMV.¹⁷ However, many argue that HT-cCMV may miss a large number of cCMV infected infants who are expected to develop hearing loss after birth.¹⁷ Lastly, delayed targeted cCMV testing can be performed later in childhood if sequelae such as SNHL develop, using banked DBS specimens. However, DBS retention laws vary by state, and such testing is subject to the lower sensitivity as mentioned earlier compared to that from saliva or urine.¹⁸

Although there have been several single center pilot studies examining cCMV screening/testing^17,19–34, to date there have been no multi-center studies comparing the positivity rates of different cCMV screening/testing approaches in the United States. Furthermore, there have been no prior multicenter studies examining factors associated with the successful implementation of these approaches. For this study, we were able to accrue factors associated with successful implementation of HT-cCMV testing. Such data are critical to inform best practices for cCMV screening, including policy, guidelines and healthcare delivery. Therefore, this study sought to fill these gaps in the literature using a multi-center cohort of infants screened for cCMV. The objectives of this study were two-fold: 1) to examine the positivity rate of cCMV screening/testing among universal, HT-cCMV and delayed targeted DBS newborn screening programs, 2) to examine the characteristics of successful HT-cCMV screening programs.

METHODS

Study overview

The NIH funded Randomized Controlled Trial of Valganciclovir for Cytomegalovirus Infected Hearing Impaired Infants (ValEAR trial ClinicalTrials.gov NCT03107871) was a randomized placebo-controlled study to determine whether valganciclovir is safe and effective in improving hearing, speech and language outcomes in cCMV infected infants with isolated SNHL. The study included 34 institutions across the US with active cCMV screening/testing programs. Sites reported their screening/testing approach (universal, HT-cCMV or delayed targeted DBS), and the number of neonates tested for cCMV. For those sites that performed HT-cCMV, the number of infants who failed their newborn hearing screening was also documented. The testing outcomes for these approaches from the existing literature were also summarized and compared to the results from this study.

Study sites

All participating sites implemented HT-cCMV (HT) and/or targeted dry blood spot testing (DBS) to diagnose cCMV as their standard of care (Table 1). The University of Minnesota also used universal cCMV screening as part of a CDC funded trial. Thirty-four sites included over 80 birth hospitals for this study. Nine sites used a targeted dried blood spot testing approach; this approach was the exclusive means for early CMV testing for one site (Seattle).

Table 1:

List of Institutions

Number	Site Name	Universal	HT-cCMV	DBS
1	Albert Einstein College of Medicine/Children’s Hospital of Montefiore (AECM)		+
2	Christiana Hospital (CCHS)		+
3	Cohen Children’s Medical Center (CCMC)		+
4	Ann and Robert H. Lurie Children’s Hospital (CHIC)		+
5	Children’s Hospital of the King’s Daughters/Eastern Virginia Medical School (CHKD)		+
6	Children’s Healthcare of Atlanta/Emory University (CHOA)		+	+
7	Columbia University Medical Center (CHON)		+	+
8	Children’s Hospital of Philadelphia/University of Pennsylvania (CHOP)		+	+
9	Connecticut Children’s Medical Center (CONN)		+
10	Children’s Hospital at Dartmouth Hitchcock (DHMC)		+
11	University of Iowa (IOWA)		+
12	Marshfield Clinic Research Institute (MCRI)		+
13	Massachusetts Eye and Ear (MEEI)		+
14	The Children’s Mercy Hospital and Clinics (MERC)		+
15	University of Michigan (MICH)		+	+
16	Medical University of South Carolina (MUSC)		+
17	University of New Mexico (NMHS)		+
18	Nationwide Children’s Hospital (NWCH)		+
19	Weill Cornell Medical Center/New York Presbyterian (NYPH)		+
20	Oregon Health and Science University (OREG)		+	+
21	Riley Hospital for Children (RILE)		+	+
22	Suny Downstate Medical Center (SDMC)		+
23	Seattle Children’s Hospital (SEAT)			+
24	Lucile Packard Children’s Hospital/Stanford (STAN)		+	+
25	SSM Health Cardinal Glennon Children’s Hospital (STLU)		+
26	Baylor College of Medicine/Texas Children’s Hospital (TCBC)		+	+
27	Rady Children’s Hospital/University of California, San Diego (UCSD)		+
28	UCSF Benioff Children’s Hospital (UCSF)		+
29	University of Massachusetts Medical School (UMAS)		+
30	University of Minnesota Masonic Children’s Hospital (UMNP)	+	+
31	Children’s Hospital of Pittsburgh (UPMC)		+
32	Primary Children’s Hospital/University of Utah (UTAH)		+
33	University of Texas Southwestern (UTSW)		+
34	Monroe Carell Jr. Children’s Hospital/Vanderbilt University Medical Center (VUMC)		+

Participants

Sixteen thousand eight hundred fifty-six infants underwent cCMV testing or screening. Eligibility for those undergoing HT-cCMV was based on failed newborn hearing screening (unilateral or bilateral) in the birth hospital. In Utah, CMV testing was performed following failed newborn hearing screening either in the birth hospital or at the first outpatient hearing screen. Universal newborn hearing screening included distortion product or transient evoked otoacoustic emissions, automated auditory brainstem response (ABR) or a combination of both.

Definitions

Diagnosis of cCMV infection consisted of a positive urine CMV PCR before 3 weeks of age. Any participant with a positive saliva PCR required a positive confirmatory urine PCR before 3 weeks of age. Infants undergoing targeted DBS testing were between 3 weeks and 12 months of age and had been diagnosed with sensorineural hearing loss identified on ABR testing, behavioral testing or had other signs or symptoms suggestive of a cCMV infection.

Sensorineural hearing loss was defined as thresholds > 20 dB at 1, 2 or 4 kHz based on behavioral responses or > 25 dB at click or any of the tone burst frequencies at 1, 2 or 4 kHz based on ABR testing. Infants with bilateral profound SNHL were excluded.

Data collection and outcomes

Study sites completed reports every other month and six times each year during the ValEAR enrollment period reporting the number of infants undergoing CMV testing and the number with a positive diagnosis. The number of infants who failed their newborn hearing screening for those performing HT-cCMV testing were also reported. The data from screening sites performing HT-cCMV testing were collected prospectively and received a central IRB approval under the University of Utah (IRB number 90760). Each participating site was added to the study via a reliance agreement. At the end of the study period, a follow up survey was sent to each site using HT-cCMV testing to obtain additional details regarding program characteristics and systems in place at each institution. The use of the survey for research purposes received an exemption under the University of Utah IRB (IRB number 112816).

Analysis

Total number of patients undergoing testing, CMV positivity rates, and SNHL prevalence were recorded for all three approaches. SNHL was diagnosed by diagnostic ABR using both air and bone conduction metrics. These values were then compared to the outcomes published in the literature. The COVID period was set as the period following March 2020 until December 2021. Analyses were completed using R version 4.1.2. We selected those sites performing HT-cCMV screening with complete data (n=336 bi-monthly records). Results from 19 sites that provided complete screening data and had an average number of failed NBHS ≥ 5 participants over 2 months were used to perform a mixed model to correlate the responses to the percent of participants who failed their NBHS and were tested for CMV. For the final model, odd ratios (OR), 95% confidence intervals (CI), and p-values were reported for each variable.

RESULTS

Thirty-four participating institutions provided CMV testing outcomes from February 2019 through December 2021. The outcomes from the 3 approaches are shown in Table 2. The positivity rate for universal cCMV screening, HT-cCMV testing and delayed DBS testing was 0.5%, 1.5% and 7.3% respectively. The largest number of infants undergoing CMV testing were in the HT-cCMV screening group (n=9017). This group also comprised the greatest number of infants detected with cCMV (n=132) for a cCMV positivity rate of 1.5%; 53 (39%) cCMV infected infants were found to have SNHL. Of the three groups, the rate of cCMV positivity was highest with DBS testing.

Table 2.

Comparing the positivity rate of three approaches for congenital cytomegalovirus.

Method	Total	cCMV+	% Positive
Universal	7670	35	0.5% (0.3,0.6)
HT-cCMV	8983	132	1.5% (1.2, 1.7)
DBS	535	39	7.3% (5.2, 9.8)

HT-cCMV = hearing targeted cytomegalovirus screening; DBS = delayed targeted dried blood spot testing, cCMV = congenital cytomegalovirus infection

Outcomes for the three approaches prior and during the COVID-19 pandemic are shown in Table 3. There were no statistically significant differences in the number of infants with failed NBHS, the number undergoing CMV testing, and the percent of those tested who were CMV positive over time. For universal CMV screening, there was a statistically significant reduction in the number tested and those who were CMV positive following the onset of the COVID-19 pandemic. For delayed DBS testing, there were no statistically significant differences in the number tested, the number who were CMV positive or the percent of those testing who were CMV positive.

Table 3.

Comparison of infants undergoing 3 different approaches before and after the onset of the COVID-19 pandemic in March 2020. Numbers are representing either average number of patients or percentage of patients (as indicated) as reported every other month.

Variable	Pre (N=117)	Post (N=226)	p-value*
Hearing Targeted: Number of patients with failed NBHS	34 (32.1)	36 (37.1)	0.63
Hearing Targeted: Number of patients tested	26.6 (28.5)	29.7 (34.3)	0.39
Hearing Targeted: Percent of failed NBHS who are tested	76.8% (22.5%)	76.3% (25.5%)	0.87
Hearing Targeted: Number of patients CMV Positive	0.4 (0.9)	0.4 (0.9)	0.98
Hearing Targeted: Percent of Tested Who Are Positive	3.9% (13.5%)	5.7% (19.7%)	0.35
Universal: Number Tested	762.3 (126.3)	387 (265.7)	0.005
Universal: CMV Positive	4.5 (1.8)	1 (0.9)	0.003
Universal: Percent of tested who are positive	0.6% (0.3%)	0.3% (0.3%)	0.072
Dry Blood Spot: Total number tested	8.7 (10.9)	7.6 (8.7)	0.70
Dry Blood Spot: Total number of CMV Positive	0.6 (1.2)	0.6 (1.1)	0.93
Dry Blood Spot: Percent of tested who are positive	8.7% (22.8%)	13.7% (31.7%)	0.47

^*All tests are t-test.

All summaries are mean (standard deviation).

To compare the programs at the hospitals utilizing HT-cCMV screening, we sent a questionnaire evaluating the practices at each institution (Supplemental Table 1) The survey was sent to 31 institutions with 26 respondents. The results are presented in Table 4. For 82% of the sites, cCMV testing after referral from newborn hearing screen was performed in the birth hospital prior to discharge. The hospitalist, neonatologist or primary care physician ordered the majority of cCMV testing. The type of testing was variable between institutions. Eighty-nine percent of institutions collected urine samples for testing and 71% of institutions collected saliva samples. This overlap in specimen type indicates that most institutions were able to perform CMV testing on either a saliva or urine sample depending on clinical context and indications for testing. Some institutions also used blood samples for cCMV testing. The primary care physician, the neonatologist and pediatric infectious disease specialists informed most families of cCMV test results.

Table 4.

Characteristics of hearing-targeted congenital CMV testing protocols among study sites (total sites surveyed n=26)

	n (%)
Testing setting
Birth hospital	23 (88%)
Outpatient clinic	10 (38%)
Conditional order set in place (vs. not)	9 (35%)
Individual(s) responsible for ordering testing *
Primary Care Physician	11 (42%)
Hospitalist/Neonatologist	19 (73%)
Nurse	5 (19%)
Other	8 (31%)
Specimen type utilized: *
Urine	23 (88%)
Saliva	18 (69%)
Dried blood spot	8 (31%)
Positive saliva results followed with urine confirmatory urine test if positive saliva	18 (69%)
Individual who informs family of diagnosis*
Primary Care Physician	16 (62%)
Neonatologist/Hospitalist	12 (46%)
Pediatric Infectious Disease	12 (46%)
Audiologist	2 (8%)
Additional work-up initiated for cCMV cases*
Complete blood count with platelets and differential	21 (81%)
Comprehensive metabolic panel	19 (73%)
Cranial ultrasound	20 (77%)
Brain MRI	7 (27%)
Diagnostic auditory brainstem response testing	26 (100%)
Referral to ophthalmology	24 (92%)
Referral to Infectious Diseases	26 (100%)
Multidisciplinary cCMV team in place	6 (23%)
Process to track cCMV patient signs/symptoms	9 (35%)
Protocol to find cCMV infants lost to follow-up	8 (31%)

^*Responses are greater than 100% because more than one answer was given in these instances.

Table 5 summarizes a mixed model analysis from 19 sites performing HT-cCMV that provided complete screening data and had an average number of failed NBHS ≥ 5 participants over two months. The two factors associated with higher proportions of cCMV testing following failed NBHS were (1) a protocol incorporating a nurse and/or a primary care physician for ordering cCMV testing, and (2) a conditional order set for cCMV testing. Institutions with additional processes in place following cCMV diagnosis were also more likely to have higher rates of cCMV testing after a failed NBHS. A few examples of these post-diagnosis processes were (1) a neonatologist or audiologist involved in informing families of a cCMV diagnosis, (2) a protocol for head ultrasound testing following a positive cCMV test, (3) a multi-institutional CMV team that evaluates patients together, and (4) a process to track CMV patients who are lost to follow-up. These factors suggested that those institutions with more extensive protocols at the point-of-care were likely to have more rigorous testing protocols in place and higher proportion of patients undergoing cCMV testing after failed NBHS.

Table 5:

Mixed models analysis correlating proportion of infants with failed NBHS who underwent CMV testing to factors involved in early CMV screening:

Protocol characteristic (present vs. not)	Odds Ratio (OR) (95% Confidence Interval)	p-value
Nurse facilitates cCMV testing	5.9 (3.4, 10.2)	< 0.001
PCP orders cCMV testing	5.4 (3.1, 9.2)	<0.001
Conditional cCMV test order set	5.3 (3.2, 8.7)	< 0.001
Provider contacting for cCMV+ test:
Neonatologist	3.7 (2.2, 6.0)	< 0.001
Primary care	0.2 (0.1, 0.3)	< 0.001
Audiologist	2.9 (1.22, 6.75)	0.015
Other	0.3 (0.2, 0.5)	< 0.001
Diagnostic testing after cCMV+ test:
Head US	20.8 (13.6, 31.5)	< 0.001
CMP	0.8 (0.6, 1.1)	0.14
Head ultrasound ordered if cCMV positive test result	20.8 (13.6, 31.5)	< 0.001
A Multidisciplinary team involved in cCMV evaluation and treatment	18.7 (6.0, 58.3)	< 0.001
Process to contact cCMV+ infants LTFU	8.4 (6.7, 10.6)	< 0.001
Track cCMV+ patient symptoms	0.26 (0.13, 0.50)	< 0.001

cCMV = congenital CMV; cCMV+ = positive congenital CMV test; Head US = head ultrasound; CMP = complete metabolic panel; LTFU = lost to follow-up

DISCUSSION

We report on the positivity rate of three different testing approaches for cCMV from over 80 birth hospitals situated throughout the US, the largest cohort study to our knowledge ever presented. Our results indicate a 0.5%, 1.5% and 7.3% rate of cCMV positivity from universal, HT-cCMV and delayed targeted DBS testing, respectively. This reported cCMV positivity rate for universal cCMV screening is comparable to a recent systematic review and meta-analysis that included 54 studies from 36 countries reporting a 0.5% prevalence (CI 0.4–0.6%) for high income countries such as for the US and Canada (Table 6).¹⁹ The results from the meta-analysis review are shown in the bottom row of the table. The rate of positivity ranged from 0.1 to 66.7%. The average positivity rate was 0.5% which is comparable to the 0.5% rate noted in our study.

Table 6:

Universal CMV Screening Studies¹

Reference	Number Tested for cCMV	Number cCMV positive (% positive)
Current Study	7670	35(0.5%)
Hildebrandt et al (1967)44	130	1 (0.8%)
Hanshaw et al (1968)45	280	6 (2.1%)
Birnbaum et al (1969)46	545	2 (0.03%)
Starr et al. (1970)47	2147	26 (1.2 %)
Melish et al (1973)48	1963	20 (1%)
Embil et al (1975)49	542	3 (0.5%)
Andersen (1979)50	3060	12 (0.4%)
Larke et al (1980)51	15,212	64 (0.4%)
Montgomery et al (1980)52	954	9 (0.9%)
Peckham et al (1983)53	14,200	42 (0.3%)
Kamada et al (1983)54	2,070	11 (0.5%)
Ahlfors et al (1984)55	10,328	50 (0.5%)
Stagno et al (1986)56	11,124	82 (0.7%)
Yow et al (1988)57	3,899	22 (0.6%)
Griffiths et al (1991)58	2,737	9 (0.3%)
Sohn et al (1992)59	514	6 (1.2%)
Fowler et al (1993)2	27,055	267 (1.0%)
Tsai et al (1996)60	1,000	18 (1.8%)
Luchsinger et al (1996)61	658	12 (1.8%)
Barbi et al (1998)62	1,268	6 (0.5%)
Boppana et al (1999)63	20,885	246 (1.2%)
Halwachs-Baumann et al (2000)64	5,867	13 (0.2%)
Schlesinger et al (2003)65	2,000	14 (0.7%)
Numazaki et al (2004)66	11,938	37 (0.3%)
Gaytant et al (2005)67	7,793	7 (0.09%)
Yamagishi et al (2006)68	1,176	2 (0.2%)
Barbi et al (2006)69	9,032	16 (0.2%)
Estripeaut et al (2007)70	317	2 (0.6%)
Foulon et al (2008)71	14,021	74(0.5%)
Engman et al (2008)72	6,060	12 (0.2%)
Endo et al (2009)73	1,010	2 (0.2%)
Boppana et al (2010)38	20,448	92 (0.5%)
Koyano et al (2011)74	21,272	66 (0.3%)
de Vries et al (2011)75	6,433	35 (0.5%)
Boppana et al (2011)76	34,989	177 (0.5%)
Paradiz et al (2012)77	2,841	4 (0.1%)
Barkai et al (2013)78	8,105	22 (0.3%)
Barkai et al (2014)79	9,845	47 (0.5%)
Waters et al (2014)80	1,044	2 (0.2%)
Pinninti et al (2015)81	73,239	266 (0.4%)
Yamaguchi et al (2017)82	23,368	60 (0.3%)
Leruez-Ville et al (2017)83	11,715	44 (0.4%)
Barlinn et al (2018)84	1,348	3 (0.2%)
Puhakka et al (2019)85	19,868	40 (0.2%)
Uchida et al (2020)86	4,125	9 (0.2%)
Yamada et al (2020)87	11,736	56 (0.5%)
Dollard et al (2021)16	12,554	56 (0.5%)
Total	450,385	2107 (0.5%)

Table modified from Figure 1 (Congenital CMV Prevalence by World Bank Income Level) Ssentongo et al. JAMA Network Open, 2021. ¹⁹

We also report on HT-cCMV testing results from a number of institutions covering a wide variety of regions across the US. In agreement to the 12 published studies using this approach, we found comparable outcomes with respect to the percentage of failed or referred newborn hearing screening and the percent tested over those who failed newborn hearing screening (Table 7). Combining the 12 studies, an average of seventy-five percent of infants who failed their newborn hearing screening underwent CMV testing. The range of testing however varied from a low of 8.5% to 100% among the different studies. The cCMV positivity rate was 2.8% with the percentage of cCMV infected infants with SNHL being 63.3%. In our study, 81% of infants who failed their newborn hearing screening underwent CMV testing. Our cCMV positivity rate was 1.5% which was lower than the cumulative rate from the existing literature. Perhaps our lower rate could be attributed to requiring a confirmatory positive urine CMV PCR due to concern for false positives with saliva CMV PCR presumably due to cross contamination from breast milk.³⁵Another factor could be the overall seropositive rate within a region which has been previously shown to effect the cCMV positivity rate.^19,34

Table 7:

Hearing Targeted Early CMV Testing Studies

Reference	Number of patients with failed NBHS	# Tested for cCMV (% = # tested/# failed NBHS)	#cCMV positive (% positive)	#SNHL/#cCMV positive (%)
Current Study	11,118	9017 (81.1%)	132 (1.5%)	53/132 (39%)
Fowler et. (2017) 17	999	999 (100%)1	31 (3.1%)	20/31 (65%)
Ari-Even et al (2016)21	200	187 (93.5%)	4 (2.3%)	3/4 (75%)
Williams et al (2014)22	411	404 (98%)2	6 (1.5%)	3/6 (50%)
Beswick et al (2019)23	283	234 (83%)	3 (1.3%)	2/3 (67%)
Diener et al (2017)24	509	234 (61.7%)	14 (6%)	4/14 (67%)
Raynor et al (2021)25	444	38 (8.5%)	2 (5.3%)	2/2 (100%)
Ronner et al (2022)26	891	530 (60%)	8 (1.5%)	3/5 (75%)3
Vancor et al (2019)31	171	171 (100%)	2 (1.2%)	1/2(50%)
Fourgeaud et al (2022)32	236	231 (98%)	2 (0.8%)	2/2 (100%)
Webb et al (2022)33	126	96 (76.2%)	1 (1%)	1/1 (100%)
Yamamoto et al (2020)34	91	91 (100%)1	7 (7.7%)	7/7 (100%)
Stehel et al (2008)88	572	483 (84%)	24 (5%)	16/24 (67%)
Total (Exclude Current Study)	4933	3698 (75%)	104 (2.8%)	64/101 (63.3%)

¹Part of a universal CMV study

²Testing rate affected by need for parental consent, having them perform the testing and returning samples

³Approximation based on data presented

The rate of cCMV infection from DBS testing reflected the yield in those infants who were found to have idiopathic SNHL or other signs and symptoms of cCMV infection and were too old to be diagnosed via urine CMV PCR testing. A review of the literature yielded four studies that reported their outcomes via a targeted delayed DBS approach (Table 8). The age distribution included infants to teenagers as old as 17 years. Forty-eight (10.1%) of the 475 participants with idiopathic SNHL were found to have cCMV infection. This cCMV positivity rate is slightly higher than the 7.3% rate noted in our study. In contrast, Choi et al. reported that 2.7% of infants who failed newborn hearing screening were found to have cCMV on DBS testing. ³⁶ This lower rate may be due to a reliance on failed hearing screening rather than SNHL as an indication for DBS testing.³⁷ Alternatively, Misono et al. found that the prevalence of cCMV infection in children with SNHL using archived DBS cards for cCMV diagnosis was 9.9% which is slightly higher than in the current study. ³⁰ Variation in positivity rates on DBS may also be related to the type of assay utilized for cCMV screening/testing with reported sensitivities ranging from 28–85.7% and specificities ranging from 72–99%.^16,38,39 The relatively high specificity of DBS PCR makes this test a good option for confirmation of cCMV infection particularly when the child is older than 3 weeks of age. However, due to the variable sensitivity of the test and the difficulty in obtaining DBS later in life, the use of DBS PCR as a testing method for cCMV may be suboptimal relative to other approaches.

Table 8:

Delayed dried blood testing studies

Reference	Age Distribution	Number Tested	Number of cCMV Positive
Current Study	1–12 months of age	535	39 (7.3%)
Pellegrinelli et al (2019)27	median 3.4 months	82	5 (6.1%)
Meyer et al (2017)28	3 mo-10 yrs	57	15 (26%)
Lee and Chan (2018)29	6 mo to 17 years	114	6 (5.3%)
Misono et al (2011)89	older than 4 yrs	222	22 (9.9%)
Total		475	48 (10.1%)

Given that the screening/testing for cCMV infection were performed before and during the COVID-19 pandemic, we attempted to determine the effect of this sentinel event on early CMV testing. For the first few months of the pandemic, there were adverse effects on newborn hearing screening, and early cCMV screening/testing.⁴⁰ We did not detect statistically significant reductions in HT-cCMV or delayed DBS testing after March 2020 suggesting the feasibility and reliability of these approaches. We did detect a reduction in the number of infants undergoing universal CMV testing and a reduction in the number who were CMV positive. However, we did not detect a statistically significant reduction in the percent positive over tested. This result contrasts with a report from Schleiss et al and Fernandez et al. citing a significant decrease in the prevalence of cCMV infection between 2019 and 2020.^41,42 The Schleiss et al. study utilized data from institutions that are included in the ValEAR study group, but also included earlier prevalence numbers for 2016–2018. We chose to evaluate all three approaches at the same time period. Thus, our universal screening outcomes may have been underpowered to detect a statistically significant reduction in positivity rates during the pandemic.

There are few studies evaluating the processes needed to establish an effective cCMV testing program. Raynor et al. identified a number of barriers to successful cCMV testing such as lack of parental awareness, discharge prior to cCMV testing, and audiology follow-up after the 3 week window needed for definitive testing.²⁵ Ronner et al. implemented their HT-cCMV program as a quality improvement project and created a multidisciplinary team that included pediatric infectious disease, otolaryngology, audiology and neonatology.²⁶ They also worked with newborn program administrators, nursing staff and laboratory personnel before starting. Lastly, they found beginning with a staged process with a few nurseries before starting a program in every nursery to be helpful. In the current study, we identified several factors associated with successful programs: a protocol that incorporates contacting cCMV positive infants who are lost to follow up, a tracking system for cCMV positive infants, cranial ultrasound following a positive cCMV test, a hospitalist involved in ordering additional testing following a positive cCMV diagnosis and a process for cCMV testing or screening. These factors suggested that those institutions with more extensive protocols at the point-of-care are likely to have more rigorous testing protocols in place and higher proportion of patients undergoing cCMV testing after failed NBHS. In contrast, factors associated with poorer outcomes consisted of relying on the primary care physician to inform the family of a positive test and those institutions utilizing blood specimens for CMV testing. Relying on already busy primary care providers to follow up on CMV testing and make appropriate referrals is likely not sufficient to ensure proper diagnosis and timely intervention. Additionally, blood samples are much more difficult to obtain than either urine or saliva and likely add to the complexity completing testing within 21 days of birth. Hopefully, identification of these positive and negative factors will aid other institutions that are planning early cCMV screening/testing to create optimal systems for implementation.

Due to the nature of screening newborns at birth hospitals, a notable limitation of these early cCMV screening/testing approaches is that there may be a number of infants inadvertently excluded. Children born at night and on weekends with relatively short hospital stays may not undergo cCMV screening/testing. Second, the outcomes of the early cCMV screening/testing approaches were highly variable amongst the sites probably due to differences in location of screening/testing, ordering protocols, notification process to families and follow-up.⁴³ One advantage for this study however is that the results reported are representative of the outcomes from real life busy clinical settings. Third, we were not able to ascertain the socioeconomic or racial makeup of these children. Thus, we cannot be certain whether this cohort is representative of a large segment of the US. However, this cohort is the largest group thus far that has undergone three different early cCMV screening or testing approaches from multiple academic institutions throughout the US.

CONCLUSION

We determined the positivity rate for cCMV from universal, HT-cCMV and delayed targeted DBS screening/testing approaches from a large US national cohort of infants. These screening/testing approaches were maintained during the COVID-19 crisis and thus demonstrate feasibility under challenging circumstances. Several factors including proactive patient follow up, specific order sets for cCMV testing, and personnel responsible for communicating with families, and protocols surrounding additional testing following a cCMV diagnosis were associated with successful HT-cCMV screening programs.

What is known on this subject:

Congenital CMV is the most common cause of non-genetic sensorineural hearing loss in children. Many infants are asymptomatic at birth but progressive sensorineural hearing loss can occur. Early diagnosis is critical in treating and possibly preventing progressive hearing loss.

What this study adds:

To date, this is the largest multi-institutional study comparing outcomes of three approaches to congenital CMV testing. The number of infants tested and the rate of congenital CMV diagnosis are evaluated. Ultimately, we demonstrate the durability of all three approaches during the COVID-19 pandemic. Factors associated with higher rates of testing for a hearing targeted approach were identified.

Abbreviations:

CMV

cytomegalovirus

cCMV

congenital cytomegalovirus

HT-cCMV

hearing targeted cytomegalovirus screening

DBS

dry blood spot