An investigation was carried out to compare the suitability of 11 open-source 3D printed visor designs, and involved feedback from a virtual focus group of healthcare workers on the suitability of the 3D printed face shield designs.

When the COVID-19 pandemic hit Ireland in spring 2020, there was a national and global shortage of personal protective equipment (PPE). This was due to the severely disrupted delivery services and supply chains, as manufacturing activity was suspended.

Additive manufacturing

Researchers at I-Form, the SFI Research Centre for Advanced Manufacturing, used their expertise in additive manufacturing – also known as 3D printing – to provide essential PPE to frontline staff, and to support other aspects of the health service.

The focus was on the printing of protective visors as shown in Figure 1. Over a period of two months, more than 4,000 visors were produced.

These were distributed to a number of hospitals including St Vincent’s, Tallaght, the Beacon and St Columcille’s; to the HSE COVID-19 test centres; and to the HSE Crisis Management Team for Dublin South, Kildare and West Wicklow.

The face shields were also sent overseas, to a healthcare centre in Bugisi, Shinyanga region in Tanzania, and knowledge and printing expertise were shared with researchers in Namibia.

A heavy-duty version of the visor, more suited to outdoor use, was produced using laser cutting; several hundred of this type of visor was supplied to ambulance services.

Figure 1: (A) Stacks of the 3D printed face shield head band (B) A single printed shield (C) The farm of printers in the I-Form lab 

The print material chosen was polylactic acid (PLA) as it is a low cost, low temperature, and a sustainable alternative to conventional polymers. The lactides are mass produced by the microbial fermentation of agricultural byproducts, mainly carbohydrate rich substances.

The head band print initially took 80 minutes per shield, by optimising the print conditions this was reduced to a mere 18 minutes. The printed shields were washed in Isopropanol and disinfected in a chlorine wash before use.

Figure 2: (A) A paramedic wearing a heavy duty laser cut face shield, (B) two staff in a local hospital wearing the 3DP face shields and (C) healthcare workers in Tanzania wearing the donated 3DP face shields 

The visor model design selected for printing was called 3DVersktan, as this happened to be the model that one of the local hospitals had begun printing and wearing at the start of the pandemic.

A range of other multiple open-source face shield designs have however, also been developed by the global 3DP community to encourage organisations and individuals to donate printed PPE to healthcare workers.

With support from SFI an investigation was carried out to compare the suitability of 11 of these open-source 3D printed visor designs. This included feedback from a virtual focus group of healthcare workers, on the suitability of the 3D printed face shield designs.

Ranking of visor performance

The investigation initially involved a ranking of visor performance based on material and design issues as detailed in Table 1. The five highest scoring face shield models were then tested to evaluate their durability.

It was noted that the scrunchie shield snapped after simple donning and doffing, while the DTM lightweight had features that would not print using a 0.8 nozzle.

These two designs were therefore eliminated, and the top three face shield designs selected for evaluation by the virtual focus group were the 3DVerkstan, the Faceshield1 and the PNSW Short (Figure 3).

Figure 3: The three printed face shields, evaluated in the survey and virtual focus group 

The visor models were given generic titles and eight healthcare workers agreed to wear each of the shields for at least one hour a day, at the same time each day and then completed a questionnaire.

Virtual focus group sessions were held with five respondents, each working for a different organisation. The results were compiled to determine what design features were most important to front-line workers and which of the three designs best met this criterion.

Five-point Likert Scale

The questionnaire was divided into three parts: the first section provided some background on the participants and their relationship with wearing PPE, using a 1-5 ranking scale, based on a five-point Likert Scale.

Participants were then asked questions specific to each face shield design, based on their experience of wearing them. Key insights collected by the survey were the ease of assembly and use, fit, comfort and robustness of the design. 

Each design was rated from 1-100 in terms of assembly, ease of use, clinical suitability, coverage, fit and whether the shield allowed enough space for other PPE.

Figure 4 details the most important design features, according to the participants, rated from 1-5. It is agreed unanimously that donning and doffing, assembly of the shield and the level of coverage that the shield provides are the most important features.

Following closely behind is the ability to sterilise them, how robust they are, the comfort levels while wearing them, their suitability for a clinical setting and whether they are certified for safe usage.

According to the survey, the less important features to the participants is the material they are manufactured from, the country where they are produced and lastly, the appearance.

Overall, the 3DVerkstan was rated the highest at 80% and the PNWS came in a close second at 76% on the 1-100 ranking in specific areas such as ease of assembly and use, see table 2.

Figure 4: Ranking of face shield parameters by medical staff based on their perceived importance 

Design recommendations

In terms of ease of assembly, ease of use and clinical suitability, the 3DVerskstan model scored the highest. The PNWS short model scored the same as the 3DVerskstan design in terms of coverage of facial area and the PNWS short scored the highest for overall fit and room for having additional PPE.

This design was larger than the 3DVerkstan model and some users preferred this. The frontline workers however disliked the open tray area of the PNWS short, claiming it was dangerous and did not give an appropriate level of coverage.

Some users also found the PLA material the face shields were printed from sore on the skin after long periods of time. Participants agreed that the acetate would need to be replaced with a more suitable and transparent polymer in the future.

It could be highly beneficial for future PPE shortages if the next generation of face shield design should be modelled from the 3DVerkstan, with the following adjustments: the tray area to be enlarged, with different sizes (small, medium and large) and printed from a softer, more flexible material for long terms of use, such as PETG.

Authors: Dr Heather O’Connor and Sylvia Leatham