To the Editor—Although being exposed to nature may accelerate healing and enhance patients’ well-being, ¹ organic material and water sources in healthcare institutions can be harmful. ^{Reference Anaissie, Penzak and Dignani2} There is little evidence on the impact of indoor gardens in healthcare settings; thus, we evaluated the potential for environmental contamination of an indoor plant wall from the time of its construction on to assess its risk for patient safety.

In 2016, an addition to our hospital was planned that included a windowless, 11-m² (118 ft²) waiting area next to the physical therapy rooms. Together with an interior designer and the infection prevention team, a “vertical garden” was conceived and installed on one of the waiting area’s walls and then assessed before the building was opened for patients. The plants (including Peperomia spp, Cryptanthus bivittatus, Pellionia repens, Philodendron scandens, Begonia rex, and Microsorum diversifolium) were inserted into a 6-m² (65ft²) polyethylenterephthalate net and irrigated with a water-loop system with a covered tank above and a water ditch below the wall. Using a microbial air sampler, particulate matter (PM-10) was measured in the surrounding air, and total aerobic bacterial counts and yeast and molds in the air were assessed using agar strips with tryptic soy and Sabouraud dextrose agar. To account for cross-contamination by other sources, we performed control measurements in the adjacent room. Aspergillus spp in the air was identified by microscopy and matrix-assisted laser desorption/ionization time-of-flight mass spectroscopy (MALDI-TOF MS). Irrigation water was tested for total aerobic bacteria, Pseudomonas spp, Legionella spp following current industry norms, and mycobacteria as described previously. ^{Reference Schreiber, Kohler, Cervera, Hasse, Sax and Keller3} Wall-surface cultures were performed using contact plates containing tryptic soy agar (for total aerobic bacteria) and dichloran glycerol and chloramphenicol (for yeast and molds). After the first results, the wall was rebuilt with a different set of Tillandsia and succulent plants and a reduced water irrigation in an attempt to minimize the contamination. Assessments were repeated as outlined above. The hospital’s infrastructure directorate approved the study and funded diagnostic testing. No institutional review board approval was required for this quality improvement project.

Measurements were performed every 14 days from 1 week before the installation of the first wall until 3 weeks after the modified wall was placed. Visual inspection revealed dust on the plant leaves and presence of flies (Drosophila spp) after 4 weeks, and biofilm formation in the water irrigation tank 6 weeks after the first wall was constructed. The results of measurements for particulate matter, bacterial and fungal burden of the surrounding air, and direct measurements on the plant wall are shown in Figure 1. We found no increases in particulate matter over time, but we encountered high concentrations in weeks 1 and 2 due to ongoing construction work in the waiting area. After the installation of the first wall, increased concentrations of total aerobic bacteria counts and yeasts and molds were detectable in the surrounding air. There was no detectable Aspergillus spp, which could be attributed to the first wall; however, we identified Aspergillus terreus in the air 2 weeks after the installation of the modified wall. Direct measurements on the wall indicated an abundance of aerobic bacteria throughout the observations. Fungi were present on plant material throughout the study period, and Aspergillus niger was isolated in 2 different samples at week 10. After the modification of the plant wall, fungal concentrations were lower than before, but total aerobic bacterial counts of the surrounding air remained elevated. In the water tank, total aerobic bacterial counts were between 150 and 3,000 CFU, with the presence of Pseudomonas spp detected in only 1 instance and no Legionella spp identified. In the draining water, neither Pseudomonas spp nor Legionella spp were present, but we identified abundant Mycobacterium avium complex, especially early after installation of the first wall.

Fig. 1. Particulate matter, total aerobic bacteria and fungi measurements of the vertical garden and its surroundings. Sample 1 was taken next to the wall, sample 2 was taken 10 m away from the wall, and the control measurement was taken in an adjacent room. Measurements denoted as location 1 were taken in the upper third of the wall, those denoted as location 2 in the middle of the wall, and those described as location 3 in the lower third of the wall.

In the present study, we evaluated an element of modern healthcare design intended to provide an enjoyable patient experience. After installing the indoor vertical garden, we detected elevated concentrations of bacteria and fungi in the air and on the wall, which could act as a source for nosocomial transmissions.

Studies on the role of plants in patient-care areas mainly assessed the impact of flowers in ornamental vases. High concentrations of gram-negative bacteria could be detected in the water inside the vases, but their relevance for nosocomial infections among immunocompetent individuals remains unclear. ^{Reference Taplin and Mertz4} Invasive fungal infections with Aspergillus spp and Fusarium spp in immunocompromised individuals have been linked to the presence of fungi in the soil of potted plants, ^{Reference Lass-Florl, Rath and Niederwieser5,Reference Rogues, Revel, Saric and Gachie6} and waterborne outbreaks due to Pseudomonas aeruginosa and Legionella pneumophila arising from water outlets and decorative fountains have been well described. ^{Reference Anaissie, Penzak and Dignani2,Reference Palmore, Stock and White7,Reference Hlady, Mullen, Mintz, Shelton, Hopkins and Daikos8}

This study has several limitations. First, we were unable to evaluate whether our findings represent an increased risk for nosocomial infections. Second, cross-contamination could have occurred due to ongoing construction works, which ended during the second wall’s assessment period. However, finalizing these construction works did not affect the measurements. Finally, the building had not dried entirely, and measurements were performed in an area that was not well ventilated, both of which might have increased the likelihood of microbial contamination. Nevertheless, relative humidity of the waiting area was within recommended norms throughout the study period.

In conclusion, the elevated concentrations of microorganisms in the air, on the plant wall, and in the water of this vertical garden led the hospital’s infection prevention committee to forego any further indoor plant installations. Building projects of indoor gardens in hospitals should be evaluated carefully to assess the potential for environmental contamination to prevent nosocomial infections.