Are we sleepwalking into PPE paralysis?

“When the facts change, I change my mind. What do you do, Sir?” – John Maynard Keynes

sleepwalkers-figure-roof-sky-royalty-free-thumbnail

Link to Dental Elf

Bottom line answer

As we approach a return to clinical practice policy makers need to be mindful that the dental profession is already highly proficient in cross infection control, and the benefits of new elaborate PPE protocols regarding aerosol generating procedures may be marginal in the light of low disease prevalence. If we need 300,000 participants in a  study it may be impossible to practically demonstrate  significant benefits to patient safety from the perfect PPE model compared to the harms created by expense and access.

 

Background

The most important element of Keynes’s famous quote relating to the current coronavirus pandemic is the word ‘fact’. Facts at this moment in time are constantly changing as this disease spreads, and what was true two weeks ago is now a distant memory. One major problem we face as far as healthcare policy is concerned is a lack of accurate base-rate data on the prevalence of the disease in the population, this was initially addresses by Professor Ioannidis (Ioannidis, 2020). His article in STAT caused quite a lot of online controversy (Bastian, 2020, Taleb, 2020) but what it did is highlight the difficulties of being objective and human. At this moment in time it is extremely difficult not to be influenced by the problems of base-rate neglect, loss aversion, and availability bias via the media as we count the daily international infection/death figures (Gaurav, 2020). In this post I want to concentrate on how important it is to understand the significance of accurate base-rate (prevalence) reporting  so we can allocate the correct amount of training and resources to the dental profession based of the potential aerosol risk in virus transmission. In a paper by Chambers on base-rates in dental decision-making there is a quote (Chambers, 1999):

‘Base your decisions on either the baseline alone or the evidence alone, depending on which one contains the most information.’

What we are seeing now is a rapid accumulation of both base-line data and evidence,  but policy decisions about the future are being based on data and precautionary principles that were only valid at the start of this pandemic. To highlight this, I would like to explore how we are going to test the real-world effectiveness of the personal protective equipment (PPE) and cross infection protocols that are flooding the profession now.

Methods

I am going to look at three areas here, base-rates (prevalence), numbers of asymptomatic individuals, and power calculations regarding PPE use. For clarity I will use natural frequencies wherever possible.

Firstly, on the 14th May the Office of National Statistics (ONS) in the UK published the results of its coronavirus (Covid-19) infection survey (ONS, 2020). This data was based on 10,705 participants’ swab tests taken over a two-week period from 27th April to 10th May, the sample was drawn from households in which someone has already participated in an ONS survey to ensure the sample was representative of the wider population. From this sample 33 individuals in 30 households tested positive for COVID-19. This equates according to the ONS to 0.27% (95% confidence interval: 0.17% to 0.41%) of the population of England.

Secondly, one of the key problems with Covid-19 is asymptomatic spread (Bai et al., 2020) but there is no reliable data so any calculations here need to be looked on as a Fermi (back of an envelope) problem. To get a best estimate on the proportion of asymptomatic patients I conducted a meta-analysis of the data presented on the Oxford Covid-19 Evidence Service (Heneghan et al., 2020). The meta-analysis was carries out in R using a random effects model (See Figure 1.)

Figure 1. Forest plot of asymptomatic individuals

Asympto

There are two points of note from this forest plot, the summary estimate for asymptomatic individuals is 27% (95% CI: 12 to 45%) and the heterogeneity between studies (variability) is extremely high.

The next stage is to put the base-rate and number of asymptomatics together in the form of a frequency tree. For ease of calculation I have rounded the figures so a base-rate of 0.27% becomes 1 in 400,  and the number of asymptomatics becomes 30% (See Figure 2.).

Figure 2. Frequency tree of asymptomatic vs symptomatic

Asympto_1

The frequency tree illustrates that in this population 1 out of every 1333 people could be an asymptomatic carrier of Covid-19.

How does this relate to dentistry? On the precautionary principle we are operating under at the moment the presumption is that all (100%) the patients are asymptomatic carriers rather than the true figure of 0.075% ( I have assumed symptomatic patients will not be attending a dental surgery or will be triaged out prior to entering the clinical environment). This becomes important when we want to test if our PPE and protocols are effectively protecting both the patients and the staff.  We now need to set up a study comparing PPE that is adequately powered to eliminate the effects of random chance around such a small prevalence statistic (Button et al., 2013). I have created three examples of PPE for aerosol generating procedures:

  • Perfect PPE model (fluid resistant disposable gowns, FFP3 masks, visors, ventilation, long fallow periods etc) with a 99%  chance of reducing viral contamination
  • Realistic expectations of enhanced PPE practice (FFP2, reusable surgical gown, rubber dam etc) at 93%.
  • Standard practice (surgical masks etc) at 80%.

I placed the data into an apriori sample size calculator (G*Power 3.19.2) with a error probability is 0.05 and power (1-b error probability) of 0.8 (See Table 1).

Table 1. Sample sizes for a well powered study into PPE effectiveness

asympto_tableAs we can see even in a simulation study, we are going to have to at least place over a hundred individuals in each arm of the study. To see if the benefit translates into the real-world, we need to go up two orders of magnitude to see if there is a significant difference between perfect and good PPE  based on accurate population base-rate figures.

Discussion

The purpose of this opinion paper was to highlight the potential problems that a precautionary principle can create in healthcare when we work on the assumption that 100% of the patients attending a dental surgery are infectious. Guidelines and protocols need to take into consideration the absolute risk within the population based on data that is accurate and up to date. Simulation studies, and pilot studies rarely carry their full reported success into the real world (Kistin and Silverstein, 2015). Without taking a deep breath and objectively assessing the changing data regarding Covid-19, policy makers, academics, and clinicians can unconsciously fall fowl of the base-rate fallacy and availability biases created by the modern media. High quality PPE and staff training is a vital component of keeping everyone safe from this virus but we must be mindful of the other effects that perfect practice can have on the health economics and affordability of health care to those most vulnerable.

Disclaimer:  The article has not been peer-reviewed; it should not replace individual clinical judgement, and the sources cited should be checked. The views expressed in this commentary represent the views of the author and not necessarily those of the host institution. The views are not a substitute for professional advice.

References

BAI, Y., YAO, L., WEI, T., TIAN, F., JIN, D. Y., CHEN, L. & WANG, M. 2020. Presumed Asymptomatic Carrier Transmission of COVID-19. JAMA.

BASTIAN, H. 2020. A rebuttal to “A fiasco in the making?” [Online]. Available: http://hildabastian.net/index.php/8-secondary/87-a-rebuttal-to-a-fiasco-in-the-making [Accessed].

BUTTON, K. S., IOANNIDIS, J. P., MOKRYSZ, C., NOSEK, B. A., FLINT, J., ROBINSON, E. S. & MUNAFÒ, M. R. 2013. Power failure: why small sample size undermines the reliability of neuroscience. Nature Reviews Neuroscience, 14, 365-376.

CHAMBERS, D. W. 1999. The roles of evidence and the baseline in dental decision making. J Am Coll Dent, 66, 60-7.

GAURAV, S. 2020. Behavioural Economics in the Fight Against COVID-19: BOMA Framework.

HENEGHAN, C., BRASSEY, C. & JEFFERSON, T. 2020. COVID-19: What proportion are asymptomatic? [Online]. Available: https://www.cebm.net/covid-19/covid-19-what-proportion-are-asymptomatic/ [Accessed].

IOANNIDIS, J. P. 2020. A fiasco in the making? As the coronavirus pandemic takes hold, we are making decisions without reliable data [Online]. STAT. Available: https://www.statnews.com/2020/03/17/a-fiasco-in-the-making-as-the-coronavirus-pandemic-takes-hold-we-are-making-decisions-without-reliable-data/ [Accessed].

KISTIN, C. & SILVERSTEIN, M. 2015. Pilot studies: a critical but potentially misused component of interventional research. Jama, 314, 1561-1562.

ONS. 2020. Coronavirus (COVID-19) Infection Survey pilot: England, 14 May 2020 [Online]. Available: https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/conditionsanddiseases/bulletins/coronaviruscovid19infectionsurveypilot/england14may2020 [Accessed].

TALEB, N. 2020. EVIDENCE BASED is often BS [Online]. Available: https://twitter.com/nntaleb/status/1240641133820207104 [Accessed].

 

Which occupations have the highest potential exposure to the coronavirus (COVID-19)?

The Office for National Statistics (ONS) has created an estimate of exposure to generic disease, and physical proximity to others, for UK occupations based on US analysis of these factors (ONS, 2020).

Dental-Hygienist

Link to The Dental Elf

Background

The ONS  produced a bubble plot on the 11th May to illustrate in their own words the:

‘clear correlation between exposure to disease, and physical proximity to others across all occupations. Healthcare workers such as nurses and dental practitioners unsurprisingly both involve being exposed to disease on a daily basis, and they require close contact with others, though during the pandemic they are more likely to be using PPE.’

From the plot it is clear to see that dentistry (dentist and dental nurse) are in the extreme top right corner denoting the highest exposure to disease and closes proximity to other people in the workplace, not only patients but also staff (See Figure 1).

Figure 1.  ONS Exposure to disease vs proximity to others

Bubble_v1

It is extremely easy to misinterpret this chart and I would argue this could be a classic case of  ‘correlation does not imply causation’.  Though technically we are as a profession, very close to our patients faces this does not imply that we, or the patients are at higher risk of catching a disease. Since the emergence of Human Immunodeficiency Virus (HIV) in the early 1980’s the dental profession has been fully aware of the risk that blood bourne (HepB+C), and respiratory infections (TB, SARS, MERS, H1N1) pose to both the patients, staff and population in general. The use of high levels of personal protective equipment (PPE), and staff who are specially trained in decontamination and cross infection measures has been normalised in the dental profession for over forty years.

Methods

To illustrate this, I thought it might be interesting to see how the dental profession compared to similar professional groups using the ONS’s ‘Occupations and exposure to disease’ and  ‘All death occurrences at ages 16 to 74 in England and Wales between 2001 and 2010’ data sets. From the 299 diagnostic codes for mortality I selected the 24 codes that represented respiratory disease excluding cancer (See Annex A). From this data five additional groups were selected where the demographics and data were comparable to dentist’s socio-economically, the 10 years mortality rate and relative risk were calculated based on weighted means.

Results

There were six professional groups were dentists, doctors, pharmacists, solicitors, higher education teaching professionals, and accountants/financial managers. Broadly speaking even though the dentists are physically closest to the patient, and potentially at the greatest risk of exposure they had 3.5 times  less respiratory disease than their non-healthcare peers. Doctors were just slightly lower risk than the dentist (See Figure 1., Table 1.)

Figure 1. Respiratory disease vs profession

respvsprof

Table 1. Occupation data.

Occupation title Dentists Doctors Pharmacists Solicitors Higher education Accountants
Proximity to others(ONS units?) 97.0 89.2 72.0 34.0 50.7 45.75
Exposure to disease(ONS units?) 90.0 91.2 76.0 14.0 11.1 3.4
Total in employment 41,000 296,000 70,000 122,000 178,000 361000
Percentage female 52.6 48.9 67.6 56 46.4 44.8
Percentage aged 55+ 11.8 16.5 14.9 18.9 25.8 16.9
Percentage BAME 28.2 27.9 32.4 14.4 9.9 11.3
Respiratory disease male age 16 to 64(excl. cancer) 6 33 10 59 95 169
Respiratory disease female age 16 to 64 (excl. cancer) 1 8 8 12 17 46
Respiratory disease total age 16 to 64 (excl. cancer) 7 41 18 71 112 215
Mortality rate (MR)x10^-4 1.71 1.39 2.57 5.82 6.29 5.96
Relative risk (RR) 1.00 0.81 1.51 3.41 3.69 3.49

 Discussion

From the data presented in this opinion piece we can clearly see that the dental profession works in an environment that poses a high risk of exposure to respiratory disease. We can also see that as a profession we suffer less respiratory disease than our peers, especially those not working in the healthcare sector, and I would propose this is due to the high degree of training regarding cross-infection and careful use of PPE. It is important however to remember that this data was collected between 2001 and 2010 so it does not represent the current situation regarding Covid-19.

Disclaimer:  The article has not been peer-reviewed; it should not replace individual clinical judgement, and the sources cited should be checked. The views expressed in this commentary represent the views of the author and not necessarily those of the host institution. The views are not a substitute for professional advice.

References

ONS. 2020. Which occupations have the highest potential exposure to the coronavirus (COVID-19)?[Online].Available: https://www.ons.gov.uk/employmentandlabourmarket/peopleinwork/employmentandemployeetypes/articles/whichoccupationshavethehighestpotentialexposuretothecoronaviruscovid19/2020-05-11 [Accessed].

Appendix

Appendix A. Respiratory diseases(excluding cancer)

Respiratory Disease
Influenza
Tuberculosis
Viral pneumonia
Pneumococcal pneumonia
Other bacterial pneumonia
Other specified pneumonia
Bronchopneumonia
Unspecified lobar pneumonia
Unspecified pneumonia
Acute bronchitis/bronchiolitis
Chronic bronchitis and emphysema
Asthma
Bronchiectasis
Coal worker’s pneumoconiosis
Asbestosis
Silicosis
Other and unspecified pneumoconiosis
Pneumoconiosis with tuberculosis
Byssinosis
Airways disease due to other specific organic dust
Farmer’s lung disease
Bird fancier’s lung
Other and unspecified allergic pneumonitis
Respiratory conditions from chemical fumes
Pulmonary fibrosis

 

 

Recommendations for the re-opening of dental services: a rapid review of international sources

International_flag_globe

Link to Dental Elf

The coronavirus 2 (SARS-CoV-2 (Covid-19)) pandemic has shut, or severely restricted the provision of only but the most essential dental care globally. From the data now currently available we have past the initial peak of infection within the population but it may still take a considerable time to develop a treatment or vaccine for the virus, in the meantime we will have to learn to function with the virus in the community. The purpose of this rapid review was to scope through the current international guidelines on re-opening dental services to help policy and decision makers establish robust practical evidence-based guidelines (Cochrane, 2020).

 Methods

 Between the 2nd and 6th May a rapid review was conducted of the international guidance  for reopening dental services utilising the WHO and the Alliance for Health Policy and Systems Research approach. A grey literature search was undertaken with the assistance of the information scientists at Cochrane Oral Health. Single data extraction was performed and quality was not assessed or validated.

Results

  • The review identified 12 guidance documents from 11 countries between the 18th March to the 5th
  • Below are listed the recommendations by domain that scored >50%:
    • Practice preparation and patient considerations
      • Patient triage by telephone (92%)
      • Social distancing in the waiting area reception (75%)
    • PPE for dental practice personnel
      • Always wear face mask (67%)
      • Unsuspected COVID-19 patients
        • Eye protection (100%)
        • FFP2 mask (50%)
      • Unsuspected COVID-19 patients undergoing AGPs
        • Disposable surgical gown (75%)
        • FFP2 mask (67%)
        • Surgical hat (50%)
      • Confirmed COVID19 patients
        • Eye protection, single use of gloves and disposable surgical gown (100%)
        • FFP2 mask (75%)
        • Surgical hat (67%)
      • Confirmed COVID19 patients undergoing AGPs
        • FFP2 mask (83%)
      • Management of the clinical room
        • Clinical room should be kept clear (50%)
      • Dental procedures.
        • Reduce or avoid AGPs (100%)
        • Reduce the risk of transmission (92%)
        • Rubber dam and high-volume suction (83%)
        • High volume suction (92%)
        • Minimally invasive procedures (50%)
      • Post-operative cleaning/disinfection/waste management
        • Cleaning and disinfection of all surfaces following every patient contact (75%)

Conclusions

The authors concluded:

‘This rapid review has provided a summary of the international guidance documents published to date. It summarises the main elements of the included documents and highlights several key messages intended to assist policy and decision makers to produce comprehensive national guidance for their own settings. In the majority of the sources addressing specific COVID-19 concerns, there was no referenced, underpinning evidence’.

 Comments

This rapid review generated a substantial volume of data that can be challenging to interpret. To help develop some context of the balance of information I extracted data both at the domain and subgroup level to produce a summary weight of consensus per domain (See Figure 1.).

RR_international

What the chart shows us is that there is a high degree of agreement about the clinical aspects of managing individual dental patients, as seen from the PPE and procedures domains, even though there is weak evidence supporting these recommendations. Guidance on how we are going to solve the re-opening problem is significantly weaker (p=0.01) in the domains of practice preparation and patient considerations, management of the clinical room, and post-operative procedures. There is nothing new here in terms of guideline construction as they are generally formulated with very little stakeholder involvement, evidence-base, or applicability (Domains 2,3 and 5 of the AGREE II criteria) (Howe, 2017). The purpose of the review was to identify these weaknesses and address them in the future guideline  development.

Disclaimer:  The article has not been peer-reviewed; it should not replace individual clinical judgement, and the sources cited should be checked. The views expressed in this commentary represent the views of the author and not necessarily those of the host institution. The views are not a substitute for professional advice.

References

COCHRANE. 2020. Recommendations for the re-opening of dental services: a rapid review of international sources [Online]. Available: https://oralhealth.cochrane.org/news/recommendations-re-opening-dental-services-rapid-review-international-sources [Accessed].

HOWE, M. S. 2017. What is the methodological quality of published dental implant guidelines? Evid Based Dent, 18, 35-36.