J. Dent. 1991; 19: 290-295

290

of

influence

biofiims

contamination R. L. S. Whitehouse, Alberta,

microbiai

on

in dental unit water

E. Peters,* J. Lizotte*

of Medical Microbiology Edmonton, Canada

Department

-a-

and C. Lilge

and Infectious

Diseases and *Department

of Oral Biology,

University

of

__--_ ~_ ABSTRACT Water from dental units (DU), used for cooling and clearing the field of dental operations, is frequently contaminated by microorganisms. Retrograde spread of oral microbes into DU tubing, contaminated plumbing systems and endogenous DU contamination have been implicated. This study investigated the contribution of DU tubing to this contamination in 11 randomly selected DU. The times required, under standardized conditions, for DU bacterial levels to decrease in response to the flushing caused by DU operation, or increase in response to stagnation caused by shutting down the DU, were measured. The DU tubing was then removed and similarly manipulated. The results showed similar bacterial levels and populations in the DU and their corresponding tubes. Sixteen control samples taken from the connecting plumbing system at distant locations, after periods of stagnation which result in DU bacterial contamination, were negative. This suggests the plumbing, in our system, is not an important factor. Thus, DU can endogenously contaminate the water passing through them; their tubes have the potential to generate similar magnitudes of bacteriai contamination to that determined from intact DU. Scanning eiectron microscopy of the tube lumens showed a biofilm, characterized by microorganisms embedded in an amorphous matrix in all cases. This biofilm could act as a reservoir to facilitate rapid recontamination. Further analysis of the data indicates there could be other contributing factors. KEY WORDS: J. Dent. 1991;

Equipment, Microbiology, Biofilms 19:

290-295

(Received 7 February 1991;

reviewed 7 March 1991;

Correspondence should be addressed to: Dr R. L. S. Whitehouse,

Department Infectious Diseases, University of Alberta, Edmonton, Alberta T6G 2H7, Canada.

INTRODUCTION high levels of microorganisms in dental unit water (DUW) are well documented (McEntegart and Clark, 1973; Gross et al., 1976; Fitzgibbon et al., 1984; Fiehn and Henriksen, 1988). These microorganisms are mostly saprophytic Gram-negative, rod-shaped bacteria (Kelstrup et al., 1977), but at least two opportunistic pathogens, Pseudomonas aeruginosa (Clark, 1974; Martin, 1987) and Legionella pneumophila (Oppenheim et al., 1987), can occur. The health hazard this represents is uncertain although oral infections attributed to I? aeruginosa exposure from DUW have been documented in immunocompromised patients (Martin, 1987). Water flushing through the dental unit (DU) tubing during DU operation results in bacterial clearance. However, the clearance is occasionally variable and always transient; subsequent DUW stagnation from nonuse of the unit results in recontamination. Attempts to control this contamination have involved the use of Persistent

Q 1991 Butterworth-Heinemann 0300-5712/91/050290-06

Ltd.

accepted 3 April 1991)

of Medical Microbiology and

various antibacterial agents. Povidone-iodine (Mills et al., 1986), Tween 80 (Kelstrup et al., 1977), chlorhexidine (Blake, 1963; McEntegart and Clark, 1973), Stericol (McEntegart and Clark, 1973), hydrogen peroxide (Kelstrup et al., 1977; Kellett and Holbrook, 1980; Exneret al., 1987) and sodium hypochlorite (Cabot Abel et al., 1970; McEntegart and Clark, 1973; Fiehn and Henriksen, 1988) have been investigated with varying degrees of success. However, the emphasis on reducing DUW bacterial levels without considering the source of the bacteria could be misdirected. Rapid recontamination following treatment of water-containing systems with antimicrobial agents is a pattern typically associated with biofilms. Biofilms are formed when bacteria in water-containing systems attach to surfaces and produce a protective polysaccharide matrix (Costerton et al., 1981,1987; Exner et al., 1987). Antibacterial agents can eradicate planktonic organisms but usually spare the bacterial populations

Whitehouse

within the film. These act as a reservoir facilitating the rapid recontamination of the water supply. This study investigated the presence of biotilms in DU tubing and their contribution to DUW bacterial contamination.

MATERIALS Bacterial

AND METHODS

contamination

in dental

units

Eleven DUs were randomly selected for investigation from the main clinic (four Adec units, Newberg, OR, USA) and pre-clinic (seven Marco units, Forest Grove, OR, USA) at the Faculty of Dentistry, University of Alberta. The pre-clinic units had not been used for treating patients. Water samples from the high speed lines were obtained by operating the DUs and collecting the effluent water aseptically in sterile test-tubes; bacterial numbers were determined during periods of use and after non-use. These were tested using the method of Miles and Misra (1938). Briefly, this involved lo-fold serial dilutions to low4 in sterile distilled water. Replicate 30 ul samples of each dilution were plated on Nutrient Agar (Tryptose Blood Agar Base with Yeast Extract, Difco Laboratories, Detroit, MI, USA) and incubated aerobically at 37°C for 48 h. Complete bacterial clearance was defined as the inability to culture organisms with this method. Preliminary experiments showed high endogenous bacterial levels in all DU; however, these were influenced by previous DU operation. Accordingly, measurements were subsequently standardized using the following protocols:

The rate at which bacteria were cleared from contaminated DUW as a result of DU operation was determined as c-11 ~~ 11r-1. n-__. __A____.--^ :-:r:_.ll_. _&_-1,._1:.__~&^ IoIIows: warer ilow rates wtxt: IIIILI~II~ SL~IIU~IUIL~~ LU 100 ml/min. Bacterial DUW contamination in each unit was lowered to undetectable levels by continuously flushing the water lines for 20 min, a time interval which preliminary studies had shown would result in complete clearance of endogenous bacterial levels. Each unit was then shut down for 48 h to permit recontamination to the original endogenous bacterial levels. The tube orifice was sealed with parafilm. Subsequently, the units were continuously operated and 2 ml water samples were collected for testing at 0.0, 0.5, 1.0, 1.5, 2.0, 3.0,4.0, 5.0 and 20.0 min.

Recontamination The rate at which DUW bacterial recontamination occurred was determined as follows: DUW bacterial contamination was brought to undetectable levels with a 20-min flush before each;ime interval for which bacterial recovery was measured. The units were shut down after each clearance for intervals of 0.0,0.5, 1.0,2.0,4.0,6.0,8.0,

291

12.0, 24.0,48.0 and 72.0 h. After each of these intervals, a 50 ml water sample was collected for testing: this volume ensured that the DUW from the internal DU tubing was sampled. As controls, 50 ml water samples were taken, after a 20min flush and a 48-h stagnation period, from taps connected to the same plumbing system and directly adjacent to the DUs; these were similarly examined. Following the same protocol, 16 further water samples were tested from faucets of the same plumbing system distant from the dental units. Bacterial contamination

of dental unit tubing

After determination of DUW contamination, each DU was shut down for at least another 48-h period before 11ll^_ ,c ,,1._._^ &L..-, .._ _,,,..,_A 11” G111“-,.&:-,” Sc;~LI”IIS“1 p”‘~u’el‘ ,a,,c; &..Li,, LU”III&._,_ we,z; IeIII”“c;u from each unit. To determine the influence of the tubing on bacterial numbers in DUW, each tube initially was tilled with sterile water which was immediately drained and tested for bacterial numbers. Subsequently, each tube was flushed (100 ml/min) for 20 min with clean tapwater and similarly tested. Then, the tubes were again filled with sterile water, sealed and maintained at room temperature for 48 h before the water was again drained and tested. As controls, the tubes were sterilized with ethylene oxide and then filled with clean DUW collected from a DU after a 20-min flush. This was drained and tested immediately. The tubes were again filled with clean DUW, maintained for a further 48 h and then tested. Statistical

Clearance

et a/.: Dental unit biofilms

analysis

The Mann-Whitney rank sums test was used to compare the bacterial counts after 48 h stagnation in the DU tubing and the 48-h counts from the clearance and recontamination experiments. Fisher’s exact test was used for proportionate comparison of resuits from samples taken from plumbing fixtures adjacent to the DUs and samples taken distant to the DU. Examination

of dental tubing

Two 5 cm segments of each tube were removed after the tubes were taken from the DUs. A smear was taken from the inside of one and stained with periodic acid Schiff (PAS) reagent; the remainingunsmearedpartwas routinely nrorec.sd r--------‘-e

fnr light __p_” mirrncrnnv __________Ti

2nd urn section rut 2nd ------ R 5 r^ ___I___-_--_---_-

similarly stained. The second segment was dehydrated through a graded series of alcohol, sputter-coated with gold and examined by SEM. As controls, new tube segments were similarly examined. Microbial

identification

After Gram staining and microscopic examination of 12 frequently occurring colonial types, characterization was attempted using the API (Analytab Products Inc.,

J. Dent.

1991;

19:

No. 5

8 6 4

0

0.5

1

1.5

2

3

4

5

20

0

0.5

1

2

Minutes

Plainsview, NY, USA) and the Scepter (Becton Dickinson Diagnostic Instrument Systems, Towson, MD, USA) systems for identifying clinical microbial isolates. Further characterization was performed at the Laboratory Centre for Disease Control (LCDC), Ottawa, Canada.

RESULTS contamination

of dental

units

Clearance Bacterial clearance occurred rapidly with continuous DU operation. All units were initially positive (mean = 1.0 X IO5colony forming units (c.f.u.)/ml, s.d. = 1.5 X 104). After 5 min, only four of 11 units were still positive (mean = 4.9 X lo2 c.f.u./ml, s.d. = 2.5 X 102);all units had cleared at 20 min (Fig. I). Recontamination Recontamination was highly variable, with the first unit testing positive again at 30 min; all units tested positive by 24 h (mean = 2.7 X lo3 c.f.u./ml, s.d. = 3.4 X 103). Bacterial counts showed further increases but were approximately similar by 48 and 72 h (Fig. 2). The hnrtwinl “II.“**Y.

6

8

12244872

Hours

Fig. 7. Bacterial clearance from DUW. Mean values from positive units are shown. The value at time 0 represents endogenous DU bacterial levels.

Bacterial

4

rmlnt frnm .,.-_ CI~Pnnit lnct -Ifor technical v.,IAA. a.“A.l _-.-w at452 . . .” h -- umc .._., -_.,. ___-_---vl-

reasons. Comparison of the 48 h bacterial count taken at the onset of the clearance experiment (1.OX 105c.f.u./ml) with 48-h count taken in the recontamination experiment (1.2 X 104c.f.u./ml) showed a statistically significant difference (P < 0.05). Control water samples from two of eight taps adjacent to the DUs were positive (mean = 1.4 X 104c.f.u./ml). All 16 water samples from taps distant from dental units were negative. This difference is not statistically significant (P > 0.05).

Fig. 2. Bacterial recontamination positive units are shown.

of DUW. Mean values from

Bacterial contamination

of dental unit tubing

Sterile water which was passed through the used DU tubing tested positive in eight cases (mean = 3.0 X lo2 c.f.u./ml), three of which had levels greater than 102c.f.u./ ml (mean = 7.2 X 102 c.f.u./ml, s.d. = 5.7 X 102). Sterile water passed through the tubes after the 20-min flush tested negative in all cases. Sterile water incubated in the used DU tubing for 48 h tested positive in all cases (mean = 1.1 X 104c.f.u./ml, s.d. = 2.5 X 102).Comparison of these 48-h tubing water samples with DUW from DUs after 48 h recontamination did not show a statistically significant difference (P > 0.05). DUW incubated in sterilized used DU tubing tested negative in all cases, immediately and after 48 h stagnation. Examination

of dental

unit tubing

Light microscopic examination of smears from the tube lumens showed masses of microorganisms in a PASpositive matrix; 5 pm sections of the same tubes showed a thin line of PAS-positive material adherent to the lumenal surface. SEM showed a variety of bacteria, including many variably sized rods, embedded in an amorphous film (Fig. 3b, c). Control tubes did not demonstrate these features (Fio 3a\. \- -D’ ---IMicrobial

identification

Similar microbial species were identified from the intact units and their tubes. The microorganisms were all aerobic Gram-negative rods. Five different species were isolated, two of which were positively identified as Pseudomonas paucimobifis and Methylobacterium mesophilica, and three tentatively identified as Neisseria sp. and a Flavoelonga ta, another Pseudomonas bacterium sp. Pseudomonas aeruginosa was not found.

Whitehouse

a

b

Fig. 3. SEM of new (a) and used (b) tubing (scale bar = 25 pm). The used tubing shows a well-developed

biofilm. Higher power SEM (c) showing a variety of rodshaped bacteria within an amorphous matrix (scale bar = 10 pm).

DISCUSSION The water supplies in DU could become contaminated in at least two ways. Contamination could occur as a result of a transient suction which develops in the water lines of a dental hand-piece when it is shut off in the mouth; this could draw oral microbes into the hand-piece and possibly the DU. This mechanism has been implicated as a cross-contamination risk as well as the major source of DU contamination (Fitzgibbon et al., 1984). However, the current importance of this bacterial source might be overemphasized. The cross-contamination risk can be minimized with the insertion of valves to prevent the suction effect and the sterilization of hand pieces (Bagga et al., 1984). Further, the suggestion that this represents the DU contaminant source is uncertain since the Gramnegative saprophytic microbes which have been consistently isolated from DUs in large quantities are general!y not found in the oral flora. In our study, no differences were noted in bacterial numbers or types between DU which had never been exposed to patients and the clinical

et al.:

Dental unit biofilms

293

DU. Finally, there is an adequate theoretical basis implicating biofilms to explain the persistent DU contamination; this represents the second and possibly more intractable problem. Biofilms are the result of the tendency of water-borne bacteria to interact with surfaces by adhesion with glycocalyx polysaccharides. A highly hydrated anionic matrix is formed which provides a particularly favourable environment for increase of bacterial numbers by cell division and recruitment. The matrix can trap nutrients and provide protection against a variety of antibacterial agents including surfactants, biocides and antibiotics (Costerton et al., 1987). Exner et al. (1982) have alluded to the possible role of the biolilm as a bacterial reservoir for replenishment of disinfected DU water supplies. More generally, they have been implicated as the focus of r-,..X-,...t ,.C-,~;,.nl IeLu11e111:,A,,,,+ III~“Ic;IIL :..c,,.+:-,I IIIIc;~LI”IIJ :, 111a” .._Aw,ue rr-;,.hr “alK;ry “1 IIIE;UI~dI appliances, such as catheters and prostheses, and as a source of fouling and material deterioration in industrial and Lappindevices (Costerton et al., 1987; Costerton Scott, 1989). With respect to the latter consideration, microbes causing material deterioration in DU would grow at room temperatures and could represent different populations than those identified in this study by culture at 37°C. This temperature was chosen to select those microbes which might be medically important. However, well-defined opportunistic pathogens, such as P. aeruginosa, identified in other studies, were not found. In this study, the presence and contribution of biofilms to DU microbial contamination was examined after developing a profile of initial DU bacterial levels and their clearance and recontamination characteristics under standardized conditions. Bacterial clearance was completed after between 0.5 and 20 min of DU operation, and recontamination occurred between 0.5 h and 24 h stagnation. Subsequent similar manipulations of the tubing removed from these units were performed at times ~_ _..-l_r- _1__..- .___,,-in __!-\ -..I_ or__. ..~ -1 ~I~ A_ cno5en to ensure Lomptere Ciedrdme (LU mm) or nedvy recontamination (48 h). Comparisons showed that bacterial numbers in the tubes and their corresponding DU after 48 h of water stagnation were similar and that the same microbial populations were cultured from the tubing and the DU. Further, biolilms could be found in all of the units and it was possible to demonstrate shedding of planktonic organisms from the tube biofilms with a single flush of sterile water in eight of the 11 cases. Thus, the similarity of the DU microbial profiles to their corresponding tubes indicates that an important, and possibly major, source of DUW bacterial contamination is the ubiquitous DU tubing biolilm. The possibility that factors, other than tubing biofilms, contribute to DU contamination cannot be excluded. The endogenous DU bacterial counts obtained at the onset of the clearance experiment were significantly higher than the bacterial counts obtained from DU at the end of the recontamination experiment. In the recontamination experiment, the DU had been subjected to repeated cycles of clearance and recontamination. It is possible that the

294

J. Dent.

1991; 19: No. 5

multiple flushes have the potential to exhaust, at least for a short period of time, the ability of the biotilm to regenerate the bacterial numbers shed into the water. Alternatively, other DU parts, such as filters or fittings, which could trap debris and contribute to the contamination, were cleared, thus reducing bacterial counts. A final consideration is the contribution of the connecting plumbing system; at least two studies have demonstrated microbial contamination directly from the connecting plumbing in large dental clinics (Fiehn and Henriksen, 1988; Kelstrup et al., 1977). Although not incompatible with these findings, Fitzgibbon et al. (1984) have presented data suggesting the major source of this contamination is exogenous to the mains water supply, a finding supported by our results. In our study, two positive water samples were collected from the attached pipes l~~rl;nn A;rm.tlxr ;ntn the ,,n;tr Tn ~nmnnvi~nn nnn~ nf the= 1t,c.&u111gUllrrLIJ lllC” Cl&r UIIIW. 111 .s”AIIycaIIu”II, ll”llU “I CI1L. sites sampled from the same plumbing system distant to the units were positive. Both sets of plumbing water samples were taken after a 48-h stagnation period (a time sufficient to regenerate a high level of contamination in DU). Further, it was not possible to generate a detectable level of contamination after 48 h by incubating clean DUW and DU tubes containing sterilized biotilms. Accordingly, we interpret the two positive results to represent retrograde spread of a primary DU contamination to the adjacent plumbing. In support of this possibility, Costerton et al. (1987) have shown that biotilms can spread up to 100 cm in as few as 3 days in biomedical systems. Presumably, in some plumbing systems, unique factors related to the plumbing construction or age, water quality and other non-specified environmental conditions could further generate or sustain biotilms to cause a more generalized contamination. These considerations indicate that attempts to deal with the problems of persistent bacterial DU contamination nl.,...lA Jll"UlU

n,.nc.,...+..n+a L"IILGIILIL1LG

n.3 "11

..,nx,n waya

fr. ..,.-_lr_ L" lEjlll""Ej

fY1.M" llll,lD

n.. -..-..A..+ "I pLG"E;IIL

their formation rather than the eradication of planktonic populations derived from the films. Further information is necessary to assess whether biotilm characteristics in the DU tubing niche have unique significance. Parameters influencing development of DU biofilms, such as conditions of water pressure, temperature or time, are not well defined. The microbial populations within the film, compared to those released into the aqueous medium, or the conditions in which different biotilm populations could be shed into the aqueous medium do not appear well understood. The possibility that a film, even if completely sterilized, could predispose to the rapid bacterial recolonization by water-borne bacteria if it is not physically removed, should be considered (Costerton and Lappin-Scott, 1989). Current stratagems to control biofilms include the industrial use of freezing cycles to produce disrupting intratilm ice crystals, or in medical devices biocideimpregnated sleeves to prevent biotilm propagation (Costerton et al., 1987). Presently, mechanical cleaning is

the simplest method for removing biotilms from DU although hydrogen peroxide preparations can cause some reduction (Exner et al., 1987). Until effective methods for controlling DU biofilm formation are developed, several chemical preparations referred to earlier appear effective for transient control of bacterial contamination.

Acknowledgements The authors wish to thank the following for their technical help with this project: MS Myrna Jackson (Charles Camsell Hospital) for bacterial identification testing, MS Pauline Ewan (Laboratory Centre for Disease Control, Ottawa) for identification of the isolates and MS Colleen Murdoch for technical assistance. We also wish to thank Dr J. W. Osborn for reviewing the manuscript and providing helpful suggestions.

References Bagga B. S. R., Murphy R. A., Anderson A W. et al. (1984) Contamination of dental unit cooling water with oral microorganisms and its prevention. J. Am. Dent. Assoc. 109, 712-716. Blake G. C. (1963) The incidence and control of bacterial infection in dental spray reservoirs. Br. Dent. .I 115, 413-416, Cabot Abel L., Miller R. L., Micik R. E. et al. (1970) Studies on dental aerobiology: IV. Bacterial contamination of water delivered by dental units. J. Dent. Res. 50, 1.567-1569. Clark A (1974) Bacterial colonization of dental units and nasal flora of dental personnel. Proc. R. Sot. Med. 67, 1269-1270. Costerton J. W. and Lappin-Scott H. M. (1989) Behaviour of bacteria in biofilms. ASM News 55,650-654. Costerton J. W., Irvin R T. and Cheng K.-J. (1981) The bacterial glycocalyx in nature and disease. Arm. Rev. Microbial. 35,299-324. Costerton J. W., Cheng K.-J., Geesey G. G. et al. (1987)

Bacteriai biofiims in nature and disease. _&in. Rev. Microbial. 41, 435-464. Exner M., Tuschewitzki G. J. and Haun F. (1982) Scanning electron microscopy of the contaminated surface of water leading plastic tubes. Zentralbl. Bakteriol. Mikrobiol. Hyg. [B] 176,425-434. Exner M., Tuschewitzki G. J. and Sharnagel J. (1987) Influence of biotilms by chemical disinfectants and mechanical cleaning. Zentralbl. Bakteriol. Mikrobiol. Hyg. [B] 183, 549-563. Fiehn N.-E. and Henriksen K. (1988) Methods of disinfection of the water system of dental units by water chlorination. J. Dent. Res. 67, 1499-1504. Fitzgibbon E. J., Bartzokas C. A., Martin M. V. et al. (1984) The source, frequency and extent of bacterial contamination of dental unit water systems. Br. Dent. J. 157, 98-101. Gross A., Devine M. J. and Cutright D. E. (1976) Microbial contamination of dental units and ultrasonic scalers. J. Periodontal. 47, 670-673. Kellett M. and Holbrook W. P. (1980) Bacterial contamination of dental handpieces. J. Dent. 8,249-253. Kelstrup J., Funder-Nielsen T. D. and Theilade J. (1977) Microbial aggregate contamination of water lines in dental equipment and its control. Acta Pathol. Microbial. Stand. [B] 85, 177-183.

Whitehouse

Martin M. V. (1987) The significance of the bacterial contamination of dental unit water systems. Br. Dent J. 163, 152-154. McEntegart M. G. and Clark A. (1973) Colonisation of dental units by water bacteria. Br. Dent J. 134, 140-142. Miles A A. and Misra S. S. (1938) The estimation of the bactericidal power of the blood. J. Hyg. (Camb.) 38, 732-749.

et al.: Dental unit biofilms

295

Mills S. E., Lauderdale P. W. and Mayhew R. B. (1986) Reduction of microbial contamination in dental units with povidone-iodone 10%..Z.Am. Dent. Assoc. 113, 280-284. Oppenheim B. A., Sefton A. M., Gill 0. N. et al. (1987) Widespread Legionella pneumophila contamination of dental stations in a dental school without apparent human infection. Epidemiol. Infect. 99, 159-l 66.

Book Review Oral Radiology, 2nd edition. H. G. Poyton and M. J. Pharaoh. Pp. 412. Decker, Toronto. Hardback, f 37.00

1989.

B. C.

This book, the authors say, is intended for students, both undergraduate and postgraduate, practising dentists, specialists, and those engaged in oral radiology. They outline their methodical approach to radiological :..+n.-...+~+;-.. ,.r-,Qn+.?,r Q.nA+h;,z ;* fn,ln,.,,& Il,L~;ll-“~‘L~““‘I :n III cl...-. III-Z Krr+ ll,J, L,#cl~L.z’ 6111” ,,,,a 10 IVII”““~” h,, ur radiation physics and radiation biology. The principles of radiographic technique take three chapters and the bulk of the book, about 20 chapters, deals with interpretation. The authors wish ‘to emphasize the importance of the radiological interpretation and the fact that other aspects should be secondary to it’ is clearly illustrated, in that they follow a pathoanatomical classification with separate chapters for, for example, infection, cysts, endocrine abnormalities and benign and malignant neoplasms: there are also chapters on the temporomandibular joint, maxillary sinus and the skull and salivary glands. The lists of references at the end of each chapter have not been updated. The book provides both student and practitioner with a wealth of radiographs illustrating and explaining different entities of the oral and dental tissues. What better basis is there for interpretation than the

opportunity to look at many such radiographs? In this case, they are sometimes paired with schematic drawings of the radiographs which certainly enhance the text, but in many cases the radiographs themselves, in particular the extraoral, are illegible. The book embodies an approach whereby diagnostic oral radiology is based on a minute examination of details of the radiographic anatomy and pathology. Infrequent or relatively infrequent diseases cam “ I” nvnmnlifim-i U,.“‘,‘r.VS”“.. in II. detail ..I.“... Unwmmr . .V..“._., anv . . .. .ctrlwtl~red ...__._.__ snnrnarh -rr’-.e-” to radiological interpretation, which is essential, especially for students, is missing. Further, the new imaging modalities mean that a textbook on the subject should also consider how these modify the approach to diagnostic problems. The process of diagnosis is a subtle interplay of several trains of thought and observation in which radiographic examination is frequently used to check the hypothesis. It, therefore, provides one step on the diagnostic pathway and should interact with the outcome of other investigations. The knowledge of how to evaluate clinical information is essential for the design of sensible diagnostic pathways. We need more, not less, to get the structure right, a challenge to us as oral radioiogists, the jacks of all clinical trades, whether we are acting as clinicians, teachers or, in particular, textbook writers. M. Rohlin