Health Hazards in the Home Environment
A Risk Assessment Methodology

[Paper presented at the Healthy Building 2006 conference, Lisbon]

1 Introduction

A risk assessment methodology has been developed that grades the severity of the estimated threat to health and safety from defects and deficiencies in housing.  It provides a means for comparing hazards in different dwellings, and for comparing widely differing hazards.  In England, this system has been adopted as the statutorily prescribed method for assessing housing conditions as the first stage in determining whether enforcement action should be taken.  However, it may be that the original development and the principles could be adapted and developed for assessing threats from conditions in the housing environment, for profiling conditions from housing survey data, and perhaps other uses.

The chance to develop the system, the Housing Health and Safety Rating System (HHSRS), arose through an opportunity to review the statutory minimum standard for housing and the principles behind it.  This influenced the approach and the development of the HHSRS.

1.1 Background to the English Housing Standards

There has always been some health based justification behind the English housing standards, although by the 1990s this had become obscured.  The early 19^th century English movement for a recognition of the association between ill health and the physical environment created pressure for some form of state intervention to ensure perceived causes of disease were eliminated.  This movement promoted what became known as the ‘Sanitary Idea’, built on a notion that smells were linked with disease, that diseases were spread by miasma.  The theory went, remove the sources of smells – the rubbish, the raw sewage – and make provision for fresh air and light, and the incidence of diseases would be reduced.

In 1868 Parliament placed duties on local authorities to identify and deal with dwellings that were ‘in a Condition or State dangerous to Health so as to be unfit for Human Habitation’ [1].  However, it wasn’t until 1919 that there was an attempt to define what was meant by the term unfit, or fit, for human habitation.  It was then that the Ministry of Health proposed a definition [2], recommending that a house should be regarded as unfit if it was not (i) free from serious dampness; (ii) satisfactorily lighted and ventilated; (iii) properly drained and provided with adequate sanitary conveniences and with a sink and suitable arrangements for the disposal of waste water; (iv) in good general repair, and had not (v) a satisfactory water supply; (vi) adequate washing accommodation; (vii) adequate facilities for preparing and cooking food; (viii) a well ventilated store.

Even the idea of a common national definition wasn’t adopted until some 35 years later when a statutory standard was introduced by the Housing Repairs and Rents Act 1954 [3].  This stated that a dwelling would be deemed unfit if it was so far defective in one or more of eight requirements as to be not reasonably suitable for occupation.  These requirements were really only general headings of matters to be taken into account, and was more limited than the 1919 recommendations.  It included repair, dampness, water supply, sanitary conveniences and cooking facilities.  However, there was no reference to washing accommodation, and it limited the requirement for lighting to ‘natural’ lighting. 

With only two minor changes – the removal of the requirement for food storage facilities, and the additional of a requirement dealing with internal arrangement [4] – this remained the national standard for the next 36 years.  It was replaced in 1990 [5] by a the new Fitness Standard that was really just an up-dated version, although it did introduce for the first time requirements for the provision of artificial lighting, for heating, for hot water, and for personal washing facilities.

1.2 The Link between Housing Conditions and Health

The original concept of what was the minimum necessary for a dwelling to be fit for its purpose – that of providing a safe and healthy shelter for the occupiers – could be said to be grounded in some notion of a link between conditions and health, albeit the Sanitary Idea.  However, the phrasing of the statutory standards were building focussed, the emphasis being on what was necessary to achieve the result.  And although it was claimed that the approach of the new Fitness Standard introduced in 1990 was to be different, the language of the Standard itself remained the same.  While the guidance [6] alluded to the health basis underlying the requirements, the appraisal of the dwelling was primarily an assessment of the structure and amenities of the dwelling – the assessment of defects and deficiencies.

1.3 Research, Research, and More Research

In 1993 Warwick Law School was commissioned to investigate the way that local authorities interpreted and applied the new Fitness Standard [7].  That study found that while there appeared to be a relatively uniform understanding and application of the Standard, there was scope for some amendments and clarifications.  As well as suggesting some improvements to the Guidance on certain requirements [8], the report identified matters that should be covered by a standard aimed at protecting health and safety.  These included –

• internal arrangement, or dangerous design features;
• thermal insulation;
• sound insulation;
• threats to health or safety from the immediate locality;
• fire precautions and means of escape in case of fire; and
• radon.

Other research was been carried out, by the Building Research Establishment (BRE), primarily reviewing the evidence of the relationship between building design and condition and the health and safety of users [9].  This review also involved checking the extent to which legal controls gave protection users, including controls applicable to existing dwellings [10].

In this research, the BRE provided a comprehensive review of the evidence and used a simple risk assessment approach to rank threats to health and safety.  The reports showed that the many of the most serious potential hazards that could be found in dwellings were not covered by the statutory standards, complementing and confirming the conclusions of the Monitoring the New Fitness Standard report [7].

Following these two studies, Warwick Law School was commissioned to carry out further work with three prime aims.  First, to identify and review all the legal provisions controlling minimum standards in existing housing with particular reference to such matters as anomalies, overlaps; and gaps.  Second, to investigate options to cover the omitted hazards highlighted by the previous studies.  And finally, working with the BRE, to investigate whether a risk assessment approach could be devised for housing conditions.

The report from this study proposed two options to deal with the anomalies and weaknesses identified [11].  Either minor amendments to the existing standard, or, and more fundamental, the development of a completely different approach that would allow the grading of conditions based on the severity of the threats to health and/or safety.  Adopting the more radical proposal, in July 1998, the Minister for Housing announced that a Rating System would be developed and commissioned Warwick Law School, working initially with the BRE, to do so.

2 The Development of a Hazard Rating System

The principle underlying the development work was that a dwelling, including the structure, associated outbuildings, garden, yard or other amenity space, and means of access, should provide a safe and healthy environment for the occupants and any visitors.  However, it is impossible to satisfy this ideal as some hazards are necessary and even desirable – such as electricity, gas, stairs, windows, and cooking facilities.  These necessary and unavoidable hazards should be as safe as possible.

The intention was to develop a system that –

(a) took into account both the frequency of a hazardous occurrence and the potential severity of the outcome;

(b) recognised that there are a range of possible outcomes that could result from a hazardous occurrence; and

(c) allowed comparison of the widely differing hazards that could be found in dwellings.

Taking account of both the frequency or likelihood of a hazardous occurrence and the severity of the outcome gives a truer indication of the importance of a hazard.  For example, a relatively infrequent occurrence which results in a very severe or fatal outcome may be more significant than a occurrence which happens frequently but causes a relatively minor outcome.

Any hazardous occurrence could result in one of a range of outcomes.  For example, below a window there may be soft earth on one side and spiked railings on the other; so, from a fall out of that window, there would be a 50% chance of bruising and a 50% chance of serious puncture wounds.

Figure 1 – Potential Housing Hazards adopted for the HHSRS

Figure 2 – Examples of the Classes of Harm adopted for the HHSRS

Housing hazards differ widely in their effect on health, in the potential to cause death, and in the nature of exposure. For some there needs to be a period of exposure before there is any apparent effect on health, such as excess cold and damp.  For others, the outcome can be relatively quick, such as for falls.  And while some have the potential to cause death, such as asbestos, for other, such as noise, death is very unlikely.  A simple and obvious means to allow comparison of the risk from all housing hazards would be to devise a system that generated a numerical score to reflect that risk.

2.1 Identifying Housing Hazards

The first stage was to identify all the potential hazards that could be found in dwellings.  Building on the work by the BRE [9], a review was carried out of reported research to collect and collate evidence of the relationship between housing and the health and safety of users.  From this a list of potential hazards was drawn-up.  However, for the purposes of this development, the list was limited to those hazards attributable to a greater or lesser extent to the housing design and conditions, but excluded any attributable solely to human behaviour, household equipment, furnishings and furniture (see Figure 1).

This review provided details of the potential impact of each hazard, including the nature and severity of the health outcome.  It also gave information on dwelling features and defects that may mitigate or increase the risk for each hazard.

2.2 Range of Outcomes

The health outcomes from the identified housing hazards could be physical injury, physical or mental illness, or other health conditions.  Other work by the BRE proposed a classification and weighting system for different health outcomes reflecting the degree of incapacity suffered [12].  This work proposed seven Classes, however, only the top four were serious enough for the victim to seek medical attention and so likely to be recorded.  These top four Classes were adopted for the Rating System (see Figure 2).

Subsequent, as part of updating and refining work (with the London School of Hygiene and Tropical Medicine), the list of health outcomes for the four HHSRS Classes was extended and some re-classified [13].

2.3 Producing a numerical score

Several approaches were investigated to produce a simple formula that combined the likelihood, the spread of possible outcomes and the weightings for the Classes of Harm.  After trials, a relatively simple formula was adopted.  This formula generates a single Hazard Score as the sum of the products of the weightings for each Class of Harm which could result from the particular hazard, multiplied by the

likelihood of an occurrence, and multiplied by the set of percentages showing the spread of Harms  (see Figure 3).  To make the final numbers more manageable, the original weighting given to the Classes of Harm in the BRE’s work [12] were increased in proportion.

3 Hazards in the English Housing Stock

Both to test the effectiveness of the formula and to produce detailed information on the impact on health in the English Housing stock, data on housing characteristics and condition were matched with data on health and injuries that could be related to housing.

Details on accidental (unintentional) home injuries were obtained from data collected through the Home Accident Surveillance System (HASS) [13]. As well as details of the victim, this data included information on the site of the accident, whether any dwelling feature(s) was implicated, and the nature and seriousness of the injury caused.  Information on other health conditions was obtained from Hospital Episode Statistics (HES) [14], which gave details of inpatient admissions.  These datasets were supplemented by mortality data from the Office of National Statistics [15], Home Office Fire Statistics, and General Practice Research Database.

A Housing and Population Database was created from a range of datasets to give details on housing characteristics, including house types and age.  This database also contained information on household characteristics.   The datasets used included the 1996 English House Condition Survey [16] the census, and commercial datasets such as ACORN [17].

Both the health, accident and housing datasets included postcodes, which in the England contain an average of 14 dwellings.  Because of the low number of dwellings, there is a high degree of homogeny of housing in each postcode.  Using the postcodes, the data on mortality and morbidity was matched with the Housing and Population Database and then analysed [18].

Using the HHSRS Formula, the analyses of the matched databases produced national average Hazard Scores for each of the 29 hazards.  For each hazard it was possible to generate the national average likelihood and spread of harm outcomes for –

• up to eight dwelling types (four age bands of houses, and four age bands of flats or dwellings in multi-occupied houses); and
• four age groups of victims, so highlighting if one age group (if any) was more vulnerable to that hazard than the others.

Where –
L = the Likelihood of an occurrence
O = the Outcome expressed as a percentage for each Class of Harm
S = the row product for each Class of Harm.

Figure 3 – The HHSRS Hazard Score Formula

These results provided a hazard profile for the English housing stock, allowing hazards to be ranked by order of the risk (both the likelihood and severity of outcomes) [19].

3.1 Current uses of the HHSRS

Eight years after the concept of a health and safety based grading system for housing conditions was first suggested, the HHSRS was introduced as the statutorily prescribed method for assessing housing conditions in England for the purposes of Part 1 of the Housing Act 2004 on 6^th April 2006 [20].

Assessing the condition of existing houses cannot be based on a strict quantifiable approach.  The variations in design, construction and maintenance mean that a qualitative approach is more appropriate, involving a high degree of judgment, preferably informed professional judgment.

Since 1954, the statutory housing standards have involved two stages – first a judgment as to whether there was a failure of any of the requirements, and second, whether that failure made the dwelling not reasonably suitable for occupation.  So although the first stage was building focussed, the second required some judgment of the potential effect on occupation.  The HHSRS places the emphasis firmly on the potential effects of defects.  Using the national averages as bench-marks, the stages involved are now – first whether the conditions mean that there is a greater than average risk from any hazard, second an assessment of how much more likely is a hazardous occurrence is more likely, and third, an assessment as to whether the outcomes will be more serious than the average.  Trials using the HHSRS assessment during the development showed that this approach was more informed and logical.  It also resulted in more

consistent and justifiable results than the previous approach.  This was convincing enough for the UK government to introduce the HHSRS in the legislation.

The HHSRS is also used in the English House Condition Survey.  This is used to monitor housing conditions in England and to inform national housing policies.

4 Discussion

In part, this is a story about how research directly contributed to a change in national policy and, ultimately, a change in legislation.  This started with the review of the 1990 Fitness Standard, centring on the interpretation of a legal standard and its application.  The positive reaction to that work led to involvement in the examination of the underlying basis for such a standard; ie, the health and safety justification for the requirements.  This in turn led to a wider review of housing standards and the opportunity to recommend a shift in the approach, a shift away from a pass/fail model to risk assessment system.   It also allowed a complete revision of the underlying principles, finally moving from the Victorian Sanitary Idea to a modern evidence and risk assessment based health and safety approach.  The result is that housing assessment is focussed on the use or intended use of the structure – to provide a safe and healthy home.

The HHSRS is evidence-based and is health and safety focussed.  Reported research linking conditions in the housing environment to health were used as the basis for identifying potential housing and to provide details for the hazard profiles.  Comparison of the prevalence of hazards to provide national benchmarks were provided from the statistical analyses of matched

housing and health databases.  While the assessment is qualitative, relying on informed professional judgment, it should be based on the evidence obtained from an inspection of the premises.  This evidence underpins the decisions for any enforcement action.

The HHSRS was developed for particular purposes.  However, the underlying principles, the collected evidence on the relationship between housing conditions and health, could inform other work.

References

[1]           Artizans and Labourers Dwellings Act 1868.

[2]           Manual of Unfit Houses and Unhealthy Areas, Ministry of Health, 1919.  See also Rural Housing Manual, Ministry of Health, 1938.

[3]           Housing Repairs and Rents Act 1954, s9

[4]           Housing Act 1969, s71.

[5]           By the Local Government and Housing Act 1989, Sched 9, para 83.

[6]                Department of the Environment Circular 6/90 Annex A.

[7]                Monitoring the New Fitness Standard, (1993) HMSO, London.

[8]           For example, although the new Standard had introduced a requirement for the provision of heating, the Guidance (DoE Circular 6/90, Annex A) stated that this could be satisfied by the existence of a dedicated 13 amp electric socket in the living room.  The report recommended that there should be means of producing heat.

[9]           Building regulation and health (1995) CRC, London, Building regulations and safety (1995) CRC London, and Building regulation and security (unpublished).

[10]                Although not attributed, this overview was carried out by members of Warwick Law School.

[11]                Controlling Minimum Standards in Existing Housing (1998) LRI, Coventry.

[12]         Cheyne AJT, et al, Refinement of a risk assessment procedure: numerical weighting of severity of harm and strength of evidence.  Healthy Buildings. 1997; l3:153-158, and Raw GJ, Cayless SM, Riley J, Cox s, Cheyne A. A risk assessment procedure for health and safety in buildings London: Construction Research Communications; 2000.

[13]         Home Accident Surveillance System: 21^st Annual Report, 1997 Data (1999) Department of Trade and Industry,  London.

[14]         Hospital Episode Statistics, Department of Health,  London.  (Use overseen by Security and Confidentiality Advisory Group – the Bellingham Committee.)

[15]                Mortality Data, Office of National Statistics, London, UK.  www.statistics.gov.uk/

[16]         English House Condition Survey: 1996 (1998), Department of the Environment, London.

[17]         ACORN, CACI Ltd, London.

[18]         Full details of the creation of the database, the methodology and the results of the analyses are given in Statistical Evidence to Support the Housing Health and Safety Rating System, Vols I, II, and III (2003) ODPM, London.

[19]         For details see – Housing Health and Safety Rating System – The Guidance (Version 2) (2004) ODPM, London.  http://www.odpm.gov.uk/index.asp?id=1152820 (accessed 3 November 2005). See also Project Report – Preparation of Version 2of the Housing Health and Safety Rating System (2004) ODPM, London.  http://www.odpm.gov.uk/embedded_object.asp?id=1152842 (accessed 3 November 2005).

[20]         The Housing Health and Safety Rating System (England) Regulations 2005 (SI 2005 No.3208), HMSO, London.