Assessing the risk of Legionella in the hotel industry in Malta.

Abstract

Objectives: Different methodologies have in time been undertaken in order to find a suitable method to quantify risk exposure for different sources of Legionella. This study is an attempt to trying to identify, quantify and model the potential of risks prevailing in Maltese hotels.

Methodology: A number of hotels (n = 59), representing 54 % of the total population, were chosen using stratified random sampling, to ensure proportionate representation, both for location and category. The auditing pro-forma was pre-tested. Consequently, an ad hoc score-based questionnaire was developed, to ensure the best possible data capture on the risks of exposure arising from certain practices and processes. The basic assumption was that risk potential was related to the probability of a system or management failure occurring. Likert-type rankings were used to reduce bias and subjectivity. Standard quality assurance protocols were observed. Auditors were covered by training sessions to ensure uniformity. Data compiled from the literature review was subjected to a series of statistical and mathematical computations, and risk factors and weightings were generated. Statistical tools, including Monte Carlo simulation led to the development of a Probability Events Tree and consequently of a Legionella Risk Model (LRM), which in turn generated a Total Estimated Risk Index (TERI) for each hotel. A monitoring study was conducted separately in order to test the reliability and the applicability of the model. A random sample, (n= 720) covering for six input, namely boiler and calorifier temperatures, return flow temperature and plant room; roof tank temperatures and free chlorine levels in roof tank water, as well as free chlorine levels in sentinel taps.

Results: Total Estimated Risk Indices (TERIs) obtained by the Legionella Risk Model (LRM), revealed that mean TERI was 151.11 ± 26.72, denoting that there is still much more to be done by hoteliers in order to minimise the risk of Legionellosis. 50% of the hotels (n = 30) exceeded the critical Probability Exposure Risk Index threshold (TERI= 150), whilst 30% (n = 18) of the hotels audited fell in the high risk zone. TERIs were correlated against field data Positive correlations obtained for estimated risk in relation to temperatures of cold water in roof tanks (r =+0.113, P= < 0.0001) and cold water in sentinel taps (r = +0.037, P= < 0.0001); whilst negative correlations were obtained when estimated risk was plotted against MSI ( r = - 0.199, P= < 0.0001), chlorine levels in roof tank water (r = - 0.061, P= < 0.0001), chlorine level in cold water sentinel taps (r = - 0.312, P= < 0.0001) and nevertheless in relation with the temperature of hot water in sentinel taps ( r = - 0.018, P= < 0.0001).

Conclusion: The results obtained indicate that management of risks in Maltese hotels is lacking. Recommendations all point to the adoption of more stringent monitoring and control programmes and initiatives, and that a National Legionella Prevention Policy needs to be instituted.

Keywords: Travel-associated Legionnaires' disease; Risk assessment for Legionella; Legionella risk modelling for hotels; Risk of Exposure to Legionella in Malta.