To the Editor—Infection control (IC) precautions for hospitalized adults with influenza consist of standard, contact, and droplet precautions with single rooms recommended or with patients cohorted, but guidelines for viral influenza-like illnesses (ILIs) are not standardized. ¹ During the January 2015 influenza A (H₃N₂) epidemic in our location, the high volume of patients with ILIs became problematic, creating a major strain on bed availability.

In January 2015, a total of 54 adults were admitted with influenza A (H₃N₂) and 37 adults were admitted with viral ILIs diagnosed by multiplex polymerase chain reaction (PCR) assay of nasopharyngeal swab samples. Of the 54 influenza case patients, 53 (98%) had influenza A (H₃N₂) and 1 (2%) had influenza B. One patient had a dual-positive rapid influenza diagnostic test for influenza A and influenza B, but PCR testing was positive for influenza A (H₃N₂).Reference D’Mello, Brammer and Blanton ^{2 –} Reference Gilca, Amini and Douville-Fradet ⁶

Of the 37 adults with viral ILIs, 16 (43%) had respiratory syncytial virus (RSV), 10 (27%) had rhinovirus/enterovirus (R/E), 5 (14%) had human parainfluenza virus type 3 (HPIV-3), 4 (11%) had human metapneumovirus (hMPV), and 2 (5%) had coronavirus (Table 1). Elderly patients, more commonly admitted for viral ILIs, had longer LOS for viral ILIs than for influenza. RSV patients were older (mean age, 83 years), with LOS similar to that of influenza patients. Importantly, HPIV-3 patients had the longest LOS of any viral ILI (mean, 19 days) and were more seriously ill, with 1 death due to HPIV-3 pneumonia. Two patients had co-colonization with influenza A (H₃N₂) and another respiratory virus, but not a worse prognosis.Reference Goka, Vallely and Mutton ^{4 –} Reference Gilca, Amini and Douville-Fradet ⁶ None of the 37 adults with viral ILIs had bacterial coinfection at admission or during hospitalization.

TABLE 1

Adults Admitted to Winthrop-University Hospital in January 2015 With a Noninfluenza Viral ILI and a Positive Respiratory Viral PCR

NOTE. hMPV: human metapneumovirus; HPIV-3: human parainfluenza virus type 3; ILI, influenza-like illness; LOS: length of stay in days; PCR, polymerase chain reaction; R/E: rhinovirus/enterovirus; RSV: respiratory syncytial virus.

^a Deceased during hospitalization.

Hand hygiene and respiratory hygiene (ie, cough etiquette) should be maintained for viral ILIs. The Centers for Disease Control and Prevention’s Guideline for Isolation Precautions recommends contact precautions for adults with hMPV, RSV, or HPIV-3 only if they are immunosuppressed. Droplet precautions and single rooms are recommended during hospitalization.Reference Siegel, Rhinehart and Jackson ⁷ Our viral PCR (FilmArray Respiratory Panel; BioFire Diagnostics) does not distinguish between rhinoviruses and enteroviruses (eg, EV-D86). R/E patients were placed on droplet and contact precautions. Since hMPV, RSV, and HPIV-3 have been implicated in serious nosocomial outbreaks in adults, we placed patients with hMPV, RSV, or HPIV-3 on standard, contact, and droplet precautions for cough duration and used single rooms when available.Reference Martinelli, Canuti and Farsani ^{8 –} Reference Widmer, Griffin, Zhu, Williams and Talbot ¹⁰

Early in the influenza epidemic here, influenza and ILI patients were given single rooms. Influenza patients had single room priority and cohorting was performed whenever possible. The bed situation was maximally stressed by adult admissions for influenza and ILIs. For patients with viral ILIs, cohorting was of limited value since patients with the same virus were usually not in the hospital at the same time.

During an influenza season or epidemic, influenza cocirculates with ILI viruses.Reference Cunha, Connolly and Abruzzo ^{3 –} Reference Goka, Vallely, Mutton and Klapper ⁵ In our view, because of potential nosocomial transmission, patients with hMPV, RSV, or HPIV-3 should be placed on the same IC precautions as those with influenza.

We experienced several unexpected findings during the January 2015 influenza A (H₃N₂) epidemic here. First, the number of adults with viral ILIs approached that of influenza. Second, relatively few ILIs were due to hMPV and coronavirus, which were associated with relatively short LOS (4 and 4.5 days, respectively). RSV (n=16) was the most common viral ILI with a mean patient age of 83 years and long LOS (8.1 days). Third, there was a relatively high incidence of R/E (n=10), but our PCR does not differentiate rhinoviruses from enteroviruses. Like RSV, R/E was most common in elderly patients (mean age, 72.6 years), and more importantly from an IC perspective, had a prolonged LOS (8 days) like RSV, impacting our bed utilization. Fourth, the most surprising finding was the relatively high number (5) of HPIV-3 patients in nonimmunosuppressed adults (mean age, 76.2 years). From an IC perspective HPIV-3 had an effect on bed utilization greater than what would be expected from the number of patients. The HPIV-3 LOS (19 days) exceeded that of all other ILIs. There was 1 death due to HPIV-3 pneumonia.

Influenza and viral ILIs cocirculate during the winter months (ie, influenza season). The number of ILIs (37 patients) during the 2015 influenza epidemic at our hospital approached that of influenza (54 patients). Admitted adults with influenza diagnosed by PCR were placed on standard, contact, and droplet precautions and were given single room priority. Influenza patients were cohorted whenever possible, but owing to high volume, single room availability quickly became extremely limited. Because some viral ILIs (eg, hMPV, RSV, HPIV-3) have potential for nosocomial spread and serious hospital outbreaks, we thought it prudent to put these ILI patients in single rooms.Reference Falsey, McElhaney and Beran ^{9 ,} Reference Widmer, Griffin, Zhu, Williams and Talbot ¹⁰ Since the viral etiology of ILIs was known by PCR, it was thought that cohorting would decrease the bed burden, but cohorting was of limited value because ILIs of the same type were not in hospital at the same time. Single room availability was further limited by the prolonged LOS of some ILI viruses—for example, RSV (8.1 days) and R/E (8 days). Although there were relatively few HPIV-3 cases, HPIV-3 LOS was the most protracted (19 days), with a disproportionate effect on bed availability. The five HPIV-3 patients were also the most ill, with 1 death due to HPIV-3 pneumonia.

We continued to provide single rooms for ILI patients for the first 3 weeks of January, but by week 4, bed availability became critical and we were forced to cohort ILIs of different viral etiologies as the influenza epidemic self-terminated. From an IC perspective we prefer diagnostic precision by PCR testing with ILIs. However, during influenza epidemics, knowing the specific viral ILI type may not be helpful when bed availability becomes severely limited.