Abstract
Aims
To examine the use and uptake of TOXBASE, an Internet database for point of care provision of poisons information in the United Kingdom during its first calendar year of web-based access.
Methods
Interrogation of the database software to examine: use by different types of user and geographical origin; profile of ingredient and product access; time of access to the system; profile of access to other parts of the database.
Results
Registered users of the system increased in the first full year of operation (1224 new users) and usage of the system increased to 111 410 sessions with 190 223 product monograph accesses in 2000. Major users were hospitals, in particular accident and emergency departments. NHS Direct, a public access information service staffed by nurses, also made increasing use of the system. Usage per head of population was highest in Northern Ireland and Scotland, and least in southern England. Ingredients accessed most frequently were similar in all four countries of the UK. Times of use of the system reflect clinical activity, with hospitals making many accesses during night-time hours. The most popular parts of the database other than poisons information were those dealing with childhood poisoning, information on decontamination procedures, teratology information and slang terms for drugs of abuse.
Conclusions
This Internet system has been widely used in its first full year of operation. The provision of clinically relevant, up to date, information at the point of delivery of patient care is now possible using this approach. It has wide implications for the provision of other types of therapeutic information in clinical areas. Web-based technology represents an opportunity for clinical pharmacologists to provide therapeutic information for clinical colleagues at the bedside.
Keywords: clinical toxicology, internet databases, poisons information
Introduction
Provision of information on therapeutics for front-line clinical staff is an important challenge for clinical pharmacologists. The development of web-based technologies makes this medium a potential option for providing information at the point of care. Developments in primary care have been reported [1] but the challenge is in the acute care of patients in the secondary care setting.
Since the 1960s poisons information centres have traditionally provided clinical staff with access via the telephone. They have used a variety of materials in support of information provision, including books, reviews, original articles and locally produced, commercial or governmental databases. The latter are either held on CD Rom or, increasingly, accessible from Poisons Centres via the Internet [2]. Poisons information has previously been transmitted to the enquirer by both phone and fax in the acute situation.
Within the United Kingdom a computer-based information system, TOXBASE, was developed for information delivery along phone lines in the 1980s, and used Viewdata technology as its platform support. This system has been previously reported [3], but was not widely adopted outside Scotland and Northern Ireland. In practice, because of its platform, it was not easy to edit the database. By the late 1990s the system was therefore dated and slow by modern standards.
Increasing demands for poisons information by telephone led the Directors of the UK National Poisons Information Service (NPIS) and the UK Departments of Health to evaluate alternative methods of poisons information delivery to UK hospitals and primary care physicians. It was agreed that the TOXBASE system would form a key element in changes proposed within the poisons information network, and that this would be placed on an Internet platform [4]. We have previously reported on the introduction of the system in the United Kingdom over the first few months of operation [5]. We now report on the first calendar year's activity, discuss the information available from the use of the system, and the implications for future provision of material via the Internet on therapeutics.
The database
TOXBASE (www.spib.axl.co.uk) is provided to registered users within the National Health Service free of charge. It is searchable by keyword (from 2001 sound search as well) and includes information on approximately 14 000 products. The database was originally constructed in 1981 [3] and has been maintained since then by the staff of the Edinburgh Centre of the NPIS with input from the other NPIS Centres. In addition to poison information it also contains a series of monographs on subjects of toxicological relevance, including childhood poisoning, deliberate chemical release, decontamination procedures, eye decontamination, teratology, slang names for drugs of abuse and non toxic household products. In addition health warnings of a toxicological nature can be posted on the database within 24 h of their receipt.
Outputs include structured on-screen clinical information and treatment advice, printable fact-sheets on individual poisons, together with the aforementioned monographs. Purpose-written user response facilities within the system include those on poisoning with new drugs, adverse reaction reports for veterinary medicines, and follow up of drugs and poisons of interest to the NPIS.
Methods
The software within the system's server individually documents inquiries to the database. The data collected includes the origin of the call, the time of the call, the parts of the database accessed, and individual product entries examined.
Information on the TOXBASE database can be categorized into the number of times the database is accessed (number of sessions), the individual products accessed, for example pharmaceuticals or household products, and the number of individual ingredients (or in the case of household products ingredient type) referred to. In addition accesses to the specific monographs on the database have also been examined.
Different user categories include: NPIS centres, hospital departments of various type, primary care physicians, and NHS Direct centres. These tend to access different product groups. For example NHS Direct centres are staffed by trained nurses and provide information directly to the public on health issues. We have therefore analysed usage of the database by user type, concentrating particularly on the products and parts of the database accessed. We have also examined the timing of database use by different groups. Finally we have examined use of the database by hospital accident and emergency departments in the four countries that make up the UK.
We have used population estimates provided by the UK Departments of Health to provide a denominator against which to compare use of the database for the geographical regions we have compared.
Results
In total 1224 new users were registered (Figure 1) during the year 2000. For comparison only 20 new registered users were added to the Viewdata system in 1998.
Figure 1.
Registered users at the end of each month.
The total numbers of users registered at the end of 2000 and their locations are shown in a Table 1. This also includes details of the numbers of sessions and product accesses in 2000. Table 2 shows the uptake of the database by different health regions in England and health boards in Scotland. Use of the database shows geographical variation, with least use in the south of England.
Table 1.
Total registrations at 31 December 2000 and usage during 2000.
Total users | Total user sessions | Total products accessed | Products/user | |
---|---|---|---|---|
A & E departments | 296 | 83 556 | 129 013 | 436 |
Other hospital departments | 815 | 9364 | 12 371 | 15 |
Poisons centres | 17 | 8545 | 33 294 | 1958 |
GPs | 673 | 1711 | 1871 | 3 |
NHS Direct | 24 | 7408 | 12 576 | 524 |
Others | 175 | 826 | 1098 | 6 |
Total/average | 2000 | 111 410 | 190 223 | 95 |
Table 2.
Usage of TOXBASE by region in England and Scotland.
Country | Health Board/Regional Office | Sessions | Sessions/100 000 population |
---|---|---|---|
England | Eastern | 6401 | 118 |
London | 2693 | 37 | |
North-west | 9730 | 148 | |
Northern and Yorkshire | 22 123 | 349 | |
South and West | 8210 | 124 | |
South-east | 5763 | 66 | |
Trent | 7136 | 139 | |
West Midlands | 9792 | 184 | |
Total/average | 71 848 | 144 | |
Scotland | Argyll & Clyde | 2352 | 551 |
Ayrshire & Arran | 1080 | 288 | |
Borders | 400 | 376 | |
Dumfries & Galloway | 692 | 470 | |
Fife | 1435 | 411 | |
Forth Valley | 1145 | 415 | |
Grampian | 2711 | 516 | |
Greater Glasgow | 2665 | 292 | |
Highland | 746 | 358 | |
Lanarkshire | 1997 | 356 | |
Lothian | 1150 | 149 | |
Orkney | 95 | 486 | |
Shetland | 0 | – | |
Tayside | 1624 | 417 | |
Western Isles | 50 | 179 | |
18 142 | 354 | ||
Northern Ireland | 13 725 | 811 | |
Wales | 6781 | 230 |
To illustrate usage of the database we have analysed the most frequent inquiries in 2000, and for comparison include data collected from the previous Viewdata system for the previous 5 years in Table 3. This shows clearly that inquiries about products containing paracetamol have remained the most frequent for the past 6 years. For paracetamol the numbers of inquiries in 2000 are more than twice as high as those for the previous year. Other changes are likely to reflect changes in the numbers of patients presenting with individual intoxications, thus inquiries about SSRI antidepressants are increasing whereas those for tricyclic antidepressants are decreasing.
Table 3.
Top ingredients accessed on TOXBASE in 1996–2000.
Position in year (total number of products accessed) | ||||||
---|---|---|---|---|---|---|
Ingredient name | 1996 (67 274) | 1997 (81 358*) | 1998 (71 605*) | 1999 (123 579*) | 2000 (190 223) | Number of accesses in 2000 (% of total) |
Paracetamol | 1 | 1 | 1 | 1 | 1 | 21 668 (11.4) |
Detergents/surfactants (not cationic) | 11 | 14 | 13 | 5 | 2 | 5276 (2.8) |
Salicylate | 2 | 2 | 2 | 2 | 3 | 5194 (2.7) |
Ibuprofen | 5 | 4 | 4 | 4 | 4 | 5057 (2.7) |
Codeine | 3 | 3 | 3 | 3 | 5 | 4690 (2.5) |
Zopiclone | 10 | 7 | 6 | 7 | 6 | 3804 (2.0) |
Diazepam | 8 | 6 | 5 | 6 | 7 | 3693 (1.9) |
Fluoxetine | 9 | 11 | 9 | 11 | 8 | 3198 (1.7) |
Caffeine | 6 | 8 | 8 | 9 | 9 | 3023 (1.6) |
Paroxetine | 14 | 10 | 12 | 9 | 10 | 2997 (1.6) |
Thioridazine | 15 | 12 | 11 | 12 | 11 | 2912 (1.5) |
Ethanol | 23 | 22 | 25 | 15 | 12 | 2800 (1.5) |
Petroleum distillates | 4 | 5 | 7 | 8 | 13 | 2722 (1.4) |
Dothiepin | 7 | 9 | 10 | 13 | 14 | 2710 (1.4) |
Sodium hypochlorite | 24 | 18 | 20 | 19 | 15 | 2604 (1.4) |
Ecstasy | 18 | 24 | 32 | 22 | 16 | 2598 (1.4) |
Dextropropoxyphene | 17 | 15 | 16 | 17 | 17 | 2390 (1.3) |
Amitriptyline | 16 | 16 | 15 | 14 | 18 | 2271 (1.2) |
Dihydrocodeine | 13 | 17 | 17 | 20 | 19 | 2298 (1.2) |
Menthol | 19 | 18 | 21 | 18 | 20 | 2064 (1.1) |
Estimated because 39 days (11%), 9 days (2%) and 91 days (25%) data were lost in 1997, 1998 and 1999, respectively, due to loss of computer statistics on the backup computer. This is not likely to affect rank order of accesses.
Times of access to the database vary with type of user and this is illustrated in Figure 2. NHS Direct inquiries, which arise from members of the public, are least frequent during night time, and those from general practice are highest in surgery hours. Hospital inquiries are least frequent between 5:00 am and 9:00 am but remain relatively constant for the rest of the day.
Figure 2.
(a) Time of access – hospitals, (b) time of access – NHS direct entries and (c) time of access – GPs.
Access from different user types is to different product groups and this is illustrated in Table 4 in which we show the most frequent inquiries from hospital accident and emergency departments, NHS Direct users and GPs for comparison. NHS Direct inquiries are more typical of exposures seen in the home in children, thus reflecting the different user base. Those from hospitals are more pharmaceutically based.
Table 4.
Top products accessed by types of users and percentage of total products accessed for that user type.
Product name | NHS Direct number (%) | Product name | A & E number (%) | Product name | GP number (%) |
---|---|---|---|---|---|
Paracetamol | 535 (4.3) | Paracetamol | 9024 (7.0) | Paracetamol | 76 (4.6) |
Ibuprofen | 192 (1.5) | Diazepam | 2873 (2.2) | Ecstasy | 38 (2.3) |
Bleach – liquid | 156 (1.2) | Aspirin | 2398 (1.9) | Batteries (button) | 33 (2.0) |
Karvol decongestant capsules | 128 (1.0) | Ibuprofen | 2342 (1.8) | Vipera berus (adder) | 30 (1.8) |
Cleaning liquid – household general | 127 (1.0) | Zopiclone | 2220 (1.7) | Lavatory cleaner (liquid bleach) | 24 (1.5) |
Holly | 122 (1.0) | Ecstasy | 1916 (1.5) | Snake-bite unknown type | 24 (1.5) |
Silica gel | 121 (1.0) | Dothiepin | 1797 (1.4) | Paraquat | 19 (1.2) |
Nail lacquer | 120 (1.0) | Amitriptyline | 1764 (1.4) | Cleaning liquid – household general | 17 (1.0) |
Vitamin pills | 118 (0.9) | Temazepam | 1667 (1.3) | Aspirin | 13 (0.8) |
Lavatory cleaner (liquid bleach) | 87 (0.9) | Co-proxamol | 1528 (1.2) | Ferrous sulphate | 13 (0.8) |
Table 5 shows the top ingredients accessed by A & E departments in the four countries that make up the UK. There are similarities between countries, with paracetamol, ibuprofen, codeine and aspirin common to all. In England inquiries about zopiclone exceeded diazepam. Thioridazine was only seen in the top 10 in Northern Ireland and Scotland.
Table 5.
Top 10 ingredients and number of accesses for A & E departments in the UK countries in 2000.
England | Northern Ireland | Scotland | Wales | |
---|---|---|---|---|
1 | Paracetamol | Paracetamol | Paracetamol | Paracetamol |
(14 415) | (3148) | (2960) | (1102) | |
2 | Ibuprofen | Codeine | Ibuprofen | Aspirin |
(3687) | (768) | (698) | (319) | |
3 | Codeine | Diazepam | Codeine | Ibuprofen |
(3041) | (724) | (638) | (236) | |
4 | Aspirin | Aspirin | Aspirin | Codeine |
(2957) | (450) | (622) | (229) | |
5 | Zopiclone | Ibuprofen | Diazepam | Ethanol |
(2856) | (429) | (547) | (208) | |
6 | Diazepam | Thioridazine | Fluoxetine | Caffeine* |
(2240) | (370) | (506) | (192) | |
7 | Fluoxetine | Temazepam | Paroxetine | Diazepam |
(2196) | (363) | (476) | (178) | |
8 | Paroxetine | Fluoxetine | Caffeine* | Sodium hypochlorite |
(2168) | (363) | (436) | (154) | |
9 | Caffeine* | Zopiclone | Zopiclone | Zopiclone |
(2056) | (348) | (434) | (152) | |
10 | Dothiepin | Caffeine* | Thioridazine | Methylenedioxy methamphetamine |
(2000) | (325) | (432) | (146) |
Caffeine from compound analgesics.
Table 6 lists the information other than specific products monographs that is accessed most frequently. Paediatric, eye and teratology inquiries were the most common.
Table 6.
Top information sources on TOXBASE.
Information | Number of accesses |
---|---|
Paediatric poisoning | 4375 |
Chemicals splashed or sprayed into the eye | 3955 |
The National Teratology Information Service | 3130 |
Oral activated charcoal | 2648 |
Slang names for drugs list | 1639 |
Slang terms used by addicts | 1438 |
Decontamination procedures | 1353 |
Blood alcohol graph | 1317 |
Problems with TOXBASE | 913 |
Deliberate release of toxic chemicals | 864 |
Discussion
The provision of therapeutic information via modem technology is feasible. Outstanding questions include its accessibility to busy clinical staff, its usefulness in practice and its accuracy. The information collected on its use is an indication of clinical activity, but does not represent all poisoning cases, or indicate the reason for the interaction with the system, which could be educational in some instances. The increasing use of the system, and pattern of enquiries do, however, illustrate the extensive acceptability of this approach to information provision, which appears to be of practical use and valued by its users (NPIS feedback). Further developments in operating software within hospital clinical information systems may allow an audit trial to be developed that will provide more information on the use of TOXBASE. Such audit systems already exist or are planned within NHS Direct systems.
The establishment of a database of the size of TOXBASE required a considerable initial outlay of staff time [3]. Approximately 2 years of dedicated staff time was needed initially for the authorship of the Viewdata database. It took 4 man-months to convert that Viewdata system for Internet use. At present over 50 man-hours a week are spent on the database, in maintenance, updating, new product entries, user support and usage analysis. This requires the skills of clinicians, poisons information officers and technical support staff.
The number of registered users of TOXBASE increased by more than 150% in the year 2000. TOXBASE is the primary database of the UK National Poisons Information Service and not surprisingly poisons centre users access the largest number of products per registered user. This group includes the National Poisons Information Centre in Dublin, which became a registered user while discussions were underway to provide TOXBASE to Irish users (available to Irish A & E users from March 2001).
The other high volume users are accident and emergency departments and NHS Direct centres. In Scotland and Northern Ireland almost all of the former have been regular users for some years, and this accounts for the high number of user sessions/100 000 population for these countries. Usage figures for Lothian in Scotland are lower than might be expected because the Accident and Emergency Department at the region's major teaching hospital, Lothian University Hospitals NHS Trust, uses a local intranet version of TOXBASE for which no statistics can be currently provided. Usage in Northern Ireland is particularly high reflecting local practice. The lower rates in England and Wales reflect the extent of coverage. Highest rates in England are in Northern and Yorkshire, North-west, and in the West Midlands where local NPIS centres have promoted TOXBASE usage.
The use of TOXBASE by NHS Direct centres is expected to increase even further as TOXBASE becomes more closely linked with their decision making software. Further NHS Direct centres, together with walk-in centres, are being established in England and Wales, and the Scottish help-line, NHS 24, will start operation in 2002.
During 2000 our users accessed 6832 of the approximately 14 000 individual products and substances. Some were accessed very many times and the total number of product accesses was 190 223. Many different products contain the same active ingredient, for example Panadol and Calpol both have paracetamol as their active ingredient. Table 3 shows the top 20 ingredients, or, for household products, ingredient types. 32% of TOXBASE product inquiries can be addressed by information on the features and treatment of poisoning involving 10 ingredients. 45% of inquiries are addressed by information on 20 ingredients. Thereafter the numbers rise exponentially. This information is used to ensure that information for frequent inquiries is updated and reviewed regularly, but also provides a profile of expected toxicology presentations.
Paracetamol is the most common agent accessed with almost three times as many accesses as the next most common. This is also the most frequent agent ingested in overdose in the UK [6, 7]. Analgesics as a group are well represented amongst frequent inquiries. Enquiries about the benzodiazepine diazepam and related hypnotic zopiclone are also common. Antidepressants occupy seven of the top 20 positions, with SSRIs generally occupying higher places than the older tricyclics. Only one antipsychotic, thioridazine, and one drug of abuse, ecstasy, are in this top 20 group.
The top ingredients accessed by the A & E departments of the four UK countries (Table 5) are similar, with the top seven ingredients in England, paracetamol, ibuprofen, codeine, aspirin, zopiclone, diazepam and caffeine, appearing in the top 10 for each of the other countries. Of the remainder all are found in the top 20 for the other countries except for temazepam which is only 29th in Scotland and 31st in Wales.
For the top products (rather than ingredients) accessed by different categories of users (A & E, NHS Direct and GPs) (Table 4) A & E departments make more inquiries regarding drugs (prescription, over the counter and drugs of abuse), while NHS Direct inquiries mainly involve over the counter drugs and household products. The numbers of enquiries from GPs are small and the information is probably skewed by those looking at the database for information or interest.
Of the information other than product monographs that is most commonly accessed, general information on paediatric poisoning is most popular. This is not surprising since almost 40% of telephone enquiries to NPIS centres concern children under the age of 10 years [8]. It would be expected that a similar proportion of TOXBASE enquiries would be about children. Other popular accesses include those about gut and eye decontamination, lists of drug slang terms and information from the National Teratology Information Service, which advises on poisoning in pregnancy. While these general monographs are not accessed as often as the top ingredients (Table 3, Table 6) they nevertheless had more than 1500 accesses each in the year. Two-way interaction with users by e-mail is also possible, and optimizing use of this facility is the next challenge.
Conclusions
Moving TOXBASE to the Internet has made it more readily accessible to end-users, and increased its use by a variety of different groups of health professionals. The most common poisons accessed have not changed with increased use. The time of access of different user groups reflects their patient client groups and their working patterns.
TOXBASE provides information at the point of healthcare delivery. Our data demonstrates the wide acceptability and use of the system, and indicates that provision of therapeutic information via modern technology is both feasible and practically useful. Information available to us, as system providers, gives an insight into the profile of clinical toxicological problems facing the different types of health professionals using the system. Web-based systems can be readily updated and carry early warnings on topics of clinical relevance. The challenge for their authors, and clinical pharmacologists as a group, is to maximize the huge educational potential of systems that deliver clinical treatment advice at the bedside.
Acknowledgments
We wish to acknowledge the help and collaboration of our colleagues in the NPIS without whom this work would not have been possible. The original concept was developed on Viewdata by Dr A. T. Proudfoot and supported by the Scottish Home and Health Department (now Scottish Executive Health Department).
References
- 1.Rogers J, Jain NL, Hayes GM. Evaluation of an implementation of PRODIGY phase two. Proceedings of the 1999 AMIA Annual Symposium; AMIA; Bethesda. [PMC free article] [PubMed] [Google Scholar]
- 2.Holohan TW, Humphreys CP, Johnson H, et al. Sources of information for acute poisoning in accident and emergency departments in Dublin, Ireland. J Toxicol Clin Toxicol. 2000;38:29–36. doi: 10.1081/clt-100100912. [DOI] [PubMed] [Google Scholar]
- 3.Proudfoot AT, Davidson WSM. A viewdata system for poisons information. Br Med J. 1983;286:1125–1127. doi: 10.1136/bmj.286.6359.125. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Information from the National Poisons Information Service. CMOS Update. 2000;25:3. [Google Scholar]
- 5.Bateman DN, Good AM, Laing WJ, Kelly CA TOXBASE. Poisons information on the Internet. Emerg Med J. 19:31–34. doi: 10.1136/emj.19.1.31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Thomas SHL, Bevan L, Bhattacharyya S, et al. Presentation of poisoned patients to accident and emergency departments in the North of England. Hum Exp Toxicol. 1996;15:466–470. doi: 10.1177/096032719601500602. [DOI] [PubMed] [Google Scholar]
- 7.Hawton K, Townsend E, Deeks J, et al. Effects of legislation restricting pack sizes of paracetamol and salicylate on self poisoning in the United Kingdom: before and after study. Br Med J. 2001;32:1–7. doi: 10.1136/bmj.322.7296.1203. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.NPIS. 2000. Edinburgh centre. Annual Report. http://www.show.scot.nhs.uk/spib.