Smoke alarms

Facts about smoke alarms and smoke alarm ownership

The Home Office draws upon four main sources for its statistical information:

• Statistical returns from fire services (Home Office Statistical Bulletin)
• The British Crime Survey (BCS)
• Pre- and Post-Community Fire Safety (CFS) campaign research
• The Attitude and Behaviour Monitor (research carried out for NCFSC).

From 1988 we have seen a steep increase in the percentage of households owning smoke alarms (8% in 1988). More recently, a survey for the Home Office in 1997 indicated that ownership was 79%, and the 1999 Attitude and Behaviour Monitor recently put that figure at 81%. Follow-up research is being carried out among the non-owners from this survey.

It should be noted that Home Office statistics based on returns from fire services report on those fires attended by them as an emergency. Any fires involving detectors where no emergency call was made to the service will not be recorded. Therefore figures reported may understate the effectiveness of fire alarms as it is unlikely that the fire services will be called if the detector is working correctly and an early-warning signal is given.

This information indicates an interesting number of points:

• Of all dwelling fires (including malicious) attended by services, 13% were first discovered by a smoke alarm. This figure has increased from 5,000 in 1993 to 9,300 in 1998;
• Those without a smoke alarm (which discovered the fire) are four times more likely to die in a fire than those who own alarms;
• Where a fire was discovered by a smoke alarm there was less damage and a higher percentage (70%) of fires had been discovered in less than five minutes. (Where smoke alarms did not discover a fire only around half were discovered in under five minutes.)

The apparently high level of smoke alarm ownership compared to 1988 disguises the fact that many of these alarms may not be working. In 1998 Home Office statistics show that in 28% of fires where an alarm was present, the alarm failed to operate. More than half these failures are due to missing or flat batteries.

Other sources indicate that in a third of cases where the alarm failed, the batteries had been removed because the alarm "kept going off", indicating that they were located too close to the kitchen, but for many the battery had simply run out (37% of those surveyed). There is little evidence from research to support the claim that batteries were removed for use in another application; in fact technicians advise that the type of battery used in smoke alarms (the PP3) is not used in many other applications these days. The main exception to this is the wireless burglar alarm, which is commonly available.

There are currently no national figures for ownership by ethnic grouping, although research has been done by some services. Research does indicate that ownership seems to be lowest among older people (55+), the group found to be most at risk from injury by fire. Attitudinal responses were generally cited for non-ownership, e.g., "it will never happen to me because I don't smoke/don't have children", etc.

The 2000 British Crime Survey (to report in January 2001) will give valuable information to services on ownership across socio-economic groupings. Existing information indicates that those at highest risk from fire are in multi-ethnic, low income group areas, followed by areas of predominantly council estates with the worst hardship. This information will allow more accurate evaluation and targeting of future campaign activity.

Table 1 Fires and casualties from fires in dwellings by presence and operation of smoke detectors, 1994-98 United Kingdom

Presence and operation of automatic alarm system
	Present, operated and raised the alarm	Present, operated, but did not raise the alarm	Present, but did not operate	Absent	Unspecified	Total
Fires2
1994	11,079	2,051	3,731	49,401	-	66,262
1995	11,625	1,907	4,099	49,165	1,580	68,376
1996	12,860	2,240	5,333	50,982	-	71,415
1997	13,704	2,526	5,814	50,275	-	72,318
1998	14,823	2,356	6,800	46,723	5	70,708
Fatal Casualties
1994	47	34	42	354	-	477
1995	58	28	71	379	13	549
1996	57	26	81	392	-	556
1997	45	27	84	398	-	554
1998	48	28	80	341	-	497
Non Fatal Casualties
1994	2,129	455	1,113	9,155	-	12,852
1995	2,205	473	1,185	9,159	42	13,064
1996	2,482	553	1,654	9,468	-	14,157
1997	2,835	573	1,830	9,624	-	14,862
1998	2,840	521	2,162	9,406	-	14,929

1. Including 'late' call and heat and smoke damage only incidents.
2. Figures are based on sample data weighted to the service totals.
3. The fatality figures for 1998 are likely to be revised downwards, as later information becomes available.
   p=provisional

Figure 2 Dwelling fires where smoke alarm failed to operate by reason, 1998 United Kingdom

Table 2 Fires in dwellings with a smoke detector where alarm did not operate by reason, 1994-98 United Kingdom

Reason	1994	1995	1996	1997	1998
Missing battery	1,698	1,786	2,281	2,545	2,985
Battery failure/flat	381	442	532	597	613
Other act preventing alarm from operating incl. turned off	498	551	756	860	883
Fire products did not reach detector(s)	469	685	1,008	975	1,336
Poor siting of detector(s)	223	142	138	186	226
Faulty system incl. incorrectly installed	178	190	219	321	288
Insufficient detectors	34	14	10	5	1
Other reason (not act or omission)	119	124	148	37	80
System apparently operational after fire	14	6	25	26	21
Other including not known, e.g. where system too badly damaged	116	159	217	261	367
Total	3,731	4,099	5,333	5,814	6,800