Epidemiology of antibiotic-associated colitis; isolation of Clostridium difficile from the hospital environment

Usefulness of semi-quantitative cultures in the diagnosis of Clostridium difficile associated disease

Semi-quantitative stool cultures on CCFA were compared to cytotoxic assays for the diagnosis of Clostridium difficile associated disease (CAD). There was a significant correlation between the amount of Clostridium difficile growth on CCFA, the presence of cytotoxin and a clinical diagnosis of CAD in the 541 initial stool specimens tested.

The clinical significance of antibiotic-associated pseudomembranous colitis in the 1990s

Antibiotic-associated pseudomembranous colitis is an uncommon but potentially serious adverse reaction, resulting in acute diarrhoea and characterised by colonic pseudomembranes. A direct relationship between the disease, recent antibiotic therapy and proliferation of Clostridium difficile in the colonic lumen was established in the late 1970s. It is thought that antibiotic therapy may alter the enteric flora, enabling C. difficile to proliferate and produce toxins with cytopathic (toxin B or cytotoxin) and hypersecretory (toxin A or enterotoxin) effects on the mucosa. Apart from clindamycin, the first antibiotic recognised to be clearly associated with pseudomembranous colitis, the antimicrobial agents most commonly responsible are cephalosporins and ampicillin (or amoxicillin). However, virtually all antibiotics except parenterally administered aminoglycosides can cause the disease. Vancomycin and metronidazole, 2 drugs used to treat antibiotic-associated pseudomembranous colitis, have also been reported to be responsible for the complication when used parenterally. Pseudomembranous colitis may develop after perioperative prophylactic antibiotic therapy with cephalosporins. Antibiotic-associated pseudomembranous colitis is most frequent in elderly and debilitated patients and in intensive care units. Nosocomial acquisition of C. difficile has been documented. Therefore it has been recommended that enteric isolation precautions should be taken with patients with this disease. The clinical symptoms include watery diarrhoea, abdominal cramping, and frequently fever, leucocytosis and hypoalbuminaemia. Toxic megacolon and acute peritonitis secondary to perforation of the colon are the most serious complications. The pseudomembranes are usually seen during endoscopic procedures, sigmoidoscopy or, if possible, colonoscopy; the most useful microbiological tests for confirmation of the diagnosis include cycloserine cefoxitin fructose agar (CCFA) stool cultures and stool toxin assays on tissues or by immunological techniques. However, cultures and toxin tests may be positive in patients without pseudomembranous colitis or C. difficile-associated diarrhoea. Mild cases may respond to discontinuation of the drug responsible, but therapy with an anticlostridial antibiotic is often necessary: a 10-day course of oral vancomycin, metronidazole or bacitracin should be given. Relapses are seen in 5 to 50% of patients treated. Antibiotic treatment should avoid sporulation leading to other relapses. 'Biotherapy' (lactobacilli, Saccharomyces) has also been proposed.

Prospective study of Clostridium difficile intestinal colonization and disease following single-dose antibiotic prophylaxis in surgery

A total of 108 volunteers undergoing an elective surgical procedure were randomly given a single 2-g intravenous prophylactic dose of either a cephalosporin or mezlocillin. Stool samples were cultured for Clostridium difficile the day before the operation and later on postoperative days 4, 7, and 14. C. difficile was detected in 23.0% of patients who received a cephalosporin (cefoxitin, 8.3%; cefazolin, 14.3%; cefotetan, 20.0%; ceftriaxone, 25.0%; cefoperazone, 43.7%), in 3.3% of patients given mezlocillin, and in none of 15 control volunteers given no antimicrobial agent. No patient experienced diarrhea.

Nosocomial Clostridium difficile colonisation and disease

To assess the risk of acquiring Clostridium difficile diarrhoea or colitis in patients colonised with C difficile, rectal swabs taken weekly for 9 weeks from patients with long-term (at least 7 days) hospital stays on three wards were cultured for C difficile. 60 (21%) of 282 patients were culture-positive for C difficile during their hospital stay, of whom 51 were symptom-free faecal excretors. C difficile diarrhoea developed in the other 9 patients; 2 were culture-positive for C difficile and had diarrhoea at the time of first culture, and 7 had diarrhoea or pseudomembranous colitis after 1-6 previously negative weekly rectal cultures. All patients with diarrhoea were on one ward, but symptom-free, excretors were found on all wards. HindIII chromosomal restriction endonuclease analysis (REA) of the C difficile isolates revealed 18 distinct types. All isolates from the patients with diarrhoea were one of two nearly identical REA types, B or B2. 26 of the 29 total B/B2 isolates were from patients on the same ward, which points to a nosocomial outbreak. The symptom-free excretors were not at increased risk of subsequent clinical illness.

Prospective, controlled study of vinyl glove use to interrupt Clostridium difficile nosocomial transmission

PURPOSE

Despite recognition that Clostridium difficile diarrhea/colitis is a nosocomial infection, the manner in which this organism is transmitted is still not clear. Hands of health care workers have been shown to be contaminated with C. difficile and suggested as a vehicle of transmission. Therefore, we conducted a controlled trial of the use of disposable vinyl gloves by hospital personnel for all body substance contact (prior to the institution of universal body substance precautions) to study its effect on the incidence of C. difficile disease.

PATIENTS AND METHODS

The incidence of nosocomial C. difficile diarrhea was monitored by active surveillance for six months before and after an intensive education program regarding glove use on two hospital wards. The interventions included initial and periodic in-services, posters, and placement of boxes of gloves at every patient's bedside. Two comparable wards where no special intervention was instituted served as controls.

RESULTS

A decrease in the incidence of C. difficile diarrhea from 7.7 cases/1,000 patient discharges during the six months before intervention to 1.5/1,000 during the six months of intervention on the glove wards was observed (p = 0.015). No significant change in incidence was observed on the two control wards during the same period (5.7/1,000 versus 4.2/1,000). Point prevalence of asymptomatic C. difficile carriage was also reduced significantly on the glove wards but not on the control wards after the intervention period (glove wards, 10 of 37 to four of 43, p = 0.029; control wards, five of 30 to five of 49, p = 0.19). The cost of 61,500 gloves (4,505 gloves/100 patients) used was $2,768 on the glove wards, compared with $1,895 (42,100 gloves; 3,532 gloves/100 patients) on the control wards.

CONCLUSIONS

Vinyl glove use was associated with a reduced incidence of C. difficile diarrhea and is indirect evidence for hand carriage as a means of nosocomial C. difficile spread.

Nosocomial acquisition of Clostridium difficile infection

We studied the acquisition and transmission of Clostridium difficile infection prospectively on a general medical ward by serially culturing rectal-swab specimens from 428 patients admitted over an 11-month period. Immunoblot typing was used to differentiate individual strains of C. difficile. Seven percent of the patients (29) had positive cultures at admission. Eighty-three (21 percent) of the 399 patients with negative cultures acquired C. difficile during their hospitalizations. Of these patients, 52 (63 percent) remained asymptomatic and 31 (37 percent) had diarrhea; none had colitis. Patient-to-patient transmission of C. difficile was evidenced by time-space clustering of incident cases with identical immunoblot types and by significantly more frequent and earlier acquisition of C. difficile among patients exposed to roommates with positive cultures. Of the hospital personnel caring for patients with positive cultures, 59 percent (20) had positive cultures for C. difficile from their hands. The hospital rooms occupied by symptomatic patients (49 percent) as well as those occupied by asymptomatic patients (29 percent) were frequently contaminated. Eighty-two percent of the infected cohort still had positive cultures at hospital discharge, and such patients were significantly more likely to be discharged to a long-term care facility. We conclude that nosocomial C. difficile infection, which was associated with diarrhea in about one third of cases, is frequently transmitted among hospitalized patients and that the organism is often present on the hands of hospital personnel caring for such patients. Effective preventive measures are needed to reduce nosocomial acquisition of C. difficile.

Laboratory diagnosis of Clostridium difficile-associated diarrhoea

This paper reviews the various laboratory procedures available for the isolation and identification of Clostridium difficile and the detection of toxins produced by this organism. Laboratories should be selective in determining which patients require investigation for Clostridium difficile-associated diarrhoea. Transport and storage of stool specimens at 4 degrees C is recommended when delays in processing may occur. Tissue culture techniques are still the best method for detection of cytotoxin and a variety of cell lines can be used. Other methods for detecting cytotoxin, and methods for detecting other toxins are not sufficiently developed yet to warrant introduction into diagnostic laboratories. Culture techniques remain the most sensitive for diagnosis, particularly since the development of a variety of enrichment techniques. Cycloserine cefoxitin fructose agar is still adequate, although reduced concentrations of antimicrobial agents are necessary, and improvements, such as the addition of sodium taurocholate, increase the recovery of spores. Enrichment cultures have markedly increased isolation rates for Clostridium difficile but the significance of these isolates needs to be carefully evaluated. Until simpler and more reliable tests are available in clinical laboratories for the detection of toxins, the isolation of Clostridium difficile from patients with diarrhoeal disease should be considered paramount.

Effects of Clostridium difficile toxins A and B in rabbit small and large intestine in vivo and on cultured cells in vitro

Clostridium difficile is recognized as the major cause of antibiotic-associated colitis. C. difficile produces two toxins, A (enterotoxin) and B (cytotoxin), that are implicated in the pathogenesis of the colitis. We examined the dose responses, time course, and synergism of these two toxins in ligated rabbit intestinal loops and in tissue culture. In rabbit small intestinal loops, toxin A caused histologically demonstrable intestinal tissue damage as early as 2 h. The secretory response greater than or equal to 8 h was similar to that of a cholera toxin control. The effect of toxin A on tissue damage or secretion was seen even if toxin was removed after 5 min. Purified toxin A caused significant net accumulation of sodium, chloride, potassium, and total protein and slightly increased osmolality of the fluid content at 6 h; these effects were similar to those caused by crude C. difficile culture filtrates containing toxins A and B. Crude C. difficile toxin caused fluid accumulation with a delayed time course in the rabbit large intestine, and in contrast to its effect in small intestine, crude toxin caused net accumulation of bicarbonate and increased pH. In tissue culture, toxin A caused a rounding up of CHO and T-84 colonic carcinoma cells. A monoclonal antibody (PCG-4) that has no effect on tissue culture cytotoxicity with toxins A and B completely inhibited the secretory and tissue-damaging effects in the intestine. Toxins A and B were synergistic in the gut only at high doses of toxin B (greater than or equal to 10 micrograms/ml), and they were additive in tissue culture. The cytopathic effect in tissue culture was not consistently associated with trypan blue uptake. The cytopathic effect of toxin A in tissue culture did not appear to involve inhibitable Ca2+-dependent or prostaglandin synthesis pathways or intact microfilament or microtubule function for its activity and was not inhibited by reducing or lysosomotropic agents. Our results suggest that toxins A and B have independent and distinct effects in vivo and in vitro.

Serotyping of Clostridium difficile

A total of 246 live Clostridium difficile cultures were serotyped by a slide agglutination technique. Fifteen grouping antisera were produced which serotyped 98% of the cultures (241 of 246). Our results indicated that certain serogroups may have specific pathogenicity. Strains of serogroups A, G, H, K, S1, and S4 were cytotoxigenic and were isolated mainly from adult patients with pseudomembranous colitis or antibiotic-associated diarrhea. Nontoxigenic strains of serogroups D and Cd-5 were isolated mainly from asymptomatic neonates and small children. Some cross-reactions occurred among some strains of serogroups A, Cd-5, G, and K. These strains were further examined by analysis of protein profiles and restriction endonuclease patterns to elucidate their serology. Typing of C. difficile by using slide agglutination is a simple technique suitable for routine examination. Serogrouping may be a useful epidemiological marker and could help in elucidating the medical relevance of some C. difficile isolates.

Clostridium difficile as a nosocomial pathogen

Patients admitted to a 19-bed floor with intermediate nursing care were studied for the onset of Clostridium difficile-associated diarrhoea during a six-month period (181 calendar days) in 1986-87. All admitted patients were reviewed weekly and followed after discharge from the study unit to other inpatient services. Multiple items in the environment of five patients' rooms were sampled bacteriologically for the presence of C. difficile weekly during the study period. Three of the rooms were selected for study because of a higher prevalence of C. difficile associated diarrhoea in the prior three years and two were selected because no cases had been discovered previously in these rooms ('control rooms'). Nine of 521 patients admitted to this unit developed C. difficile diarrhoea (1.73 cases/100 patients admitted) versus 0.30/100 patients admitted to all other sites in our hospital (24 of 7970 other patients). This represented respectively 3.91 cases per 1000 patient days on this floor versus 0.37 patients/1000 patient days throughout the hospital. Seven of the C. difficile diarrhoea cases were associated with stay in the C. difficile associated rooms, versus two cases in the two 'control rooms'. C. difficile was isolated from the toilet seats, bedpan hopper, night stands or food trays. Of some 1955 cultures taken, only 1.9% overall were positive for C. difficile.

Epidemiology and prevention of Clostridium difficile infections in a leukemia unit

A 29-month prospective study was carried out in a leukemia unit with the aim of investigating the epidemiology of Clostridium difficile infections and limiting their spread. Systematic cultures of stools and assays for cytotoxin were performed on patient admission and at weekly intervals, yielding 1,355 cultures and assays. The study period was divided in period A, before total unit renovation, and period B, afterwards. During period B all patient carriers of Clostridium difficile received vancomycin. A comparison of the two periods showed that the percentage of positive cultures fell from 16.6% to 3.6% and the positive toxin assays from 9.9% to 1.2%. It was concluded that colonization by Clostridium difficile can be prevented in hospital wards with generally high rates of infection by a combination of decontamination of the environment, introduction of preventive measures and treatment of Clostridium difficile carriage with vancomycin.

Restriction endonuclease analysis of nosocomial isolates of Clostridium difficile

A total of 110 clinical isolates of Clostridium difficile were analyzed by agarose gel electrophoresis by using both bacterial restriction endonuclease analysis (REA) and plasmid profiles. A total of 72 isolates were divided into 12 groups according to their REA patterns. Some 38 isolates exhibited unique patterns. Pattern A occurred in 20% of isolates. Isolates with patterns B, E, and G were cytotoxin negative. The remaining groups were cytotoxin positive. Multiple isolates obtained from two stool specimens were studied to examine the variation in REA profiles found in single specimens. In these specimens no variation in REA profiles was found. The stability of C. difficile was studied by examining sequential in vitro subcultures of a single isolate and strains isolated over a 4-month period from two long-term carriers. REA patterns were stable over time, both in vitro and in vivo. Because plasmid DNA was observed in 53% of isolates, plasmid profiles alone could not be used to study the spread of C. difficile; however, they were necessary for the interpretation of REA patterns in some instances.

Application of whole-cell DNA restriction endonuclease profiles to the epidemiology of Clostridium difficile-induced diarrhea

Two patients in one hospital room acquired pseudomembranous colitis, one shortly after the other. The DNA restriction patterns of isolates from the patients and of four isolates from the environment were indistinguishable from one another and differed from isolates of other patients. Restriction endonuclease digest analysis appears to be a useful method for studying the epidemiology of Clostridium difficile.

Nosocomial diarrhoeas in a surgical division hyperendemic for Clostridium difficile: epidemiologic aspects emerging from an analysis of clinical records

Having previously shown that Clostridium difficile was responsible for an intense and protracted endemic of nosocomial diarrhoeas in the surgical division of a Tuscan hospital, we started a retrospective analysis on all records from the affected division, to cover a period of 15 months. A statistical description is given of a large series of nosocomial diarrhoeas, as well as direct estimates of their incidence rates in selected high risk subgroups. The situation described is epidemiologically unique, and its study tells how an endemic may reach a steady state of high nosocomial morbidity.

Review of Clostridium difficile-associated diseases

Clostridium difficile has recently become recognized as an important nosocomial pathogen. This review summarizes what is known about the isolation of the organism, the spectrum of clinical disease, virulence factors, treatments, and methods of prevention. Risk factors for C. difficile disease are also discussed. The most important risk factor is the use of certain antibiotics (ampicillin, cephalosporins, and clindamycin). C. difficile is associated with 96% to 100% of cases of pseudomembraneous colitis, 60% to 75% of antibiotic-associated cases of colitis, and 11% to 33% of antibiotic-associated cases of diarrhea. Other risk factors include gastrointestinal manipulations, advanced age, female sex, inflammatory bowel disease, cancer chemotherapy, and renal disorders. Hospital outbreaks of C. difficile disease are examined. Data from nosocomial outbreaks support transmission of C. difficile by contaminated fomites and hand carriage by hospital personnel.

Diagnosis of Clostridium difficile-associated intestinal disease

Toxigenic Clostridium difficile is the major cause of antimicrobial agent-associated pseudomembranous colitis and is the etiological agent of approximately 30% of cases of nonspecific colitis and diarrhea (without colitis) induced by antimicrobial agents. In addition, C. difficile has been implicated in certain intestinal diseases not related to prior antimicrobial administration. C. difficile has been reported to be one of the most common enteropathogens isolated from stool specimens submitted to hospital laboratories. Thus, diagnosis of C. difficile-associated intestinal disease should now be routinely performed in diagnostic clinical laboratories. The diagnosis of C. difficile-associated intestinal disease relies on the demonstration of either the organism or the toxin(s) in stool specimens or antibody response in serum to the toxin(s). Several selective medium are available for the recovery of C. difficile from stool specimens. The toxin(s) of C. difficile can be demonstrated using a variety of techniques, including biological assays as well as immunological assays. This article will review the techniques currently available to aid in the diagnosis of C. difficile-associated intestinal disease.

Clostridium difficile-associated diarrhea follows perioperative prophylaxis with cefoxitin

Clostridium difficile-associated diarrhea during prolonged therapy of obstetric and gynecologic infections is known to occur with use of all classes of antibiotics except vancomycin and the aminoglycosides. We present 11 cases of C. difficile-associated diarrhea which followed a short course of perioperative prophylaxis with cefoxitin during a 1-year period. Nine of the cases of C. difficile-associated diarrhea were among 162 women who received cefoxitin perioperative prophylaxis for cesarean section or hysterectomy, but none occurred in 85 women who received one of four other antibiotics for perioperative prophylaxis (p = 0.024, Fisher's exact test). The two other occurrences of C. difficile-associated diarrhea following perioperative prophylaxis with cefoxitin were in women who underwent exploratory laparotomy. We conclude that C. difficile-associated diarrhea is related to perioperative prophylaxis with cefoxitin.

Serogrouping of Clostridium difficile strains by slide agglutination

Six different agglutinating antisera were obtained by immunizing rabbits with Formol-treated strains of Clostridium difficile. After appropriate absorption, these antisera were used to define six serogroups designated by the letters A, B, C, D, F, and G. Altogether, 315 strains of C. difficile from various origins were tested for slide agglutination by these antisera; 312 (99%) of them were agglutinated by one of these antisera. A and C were the most common serogroups. An excellent correlation, ranging from 85 to 100%, was found between the serogroup and the toxigenicity of the strains. The correlation between serogroup and sorbitol fermentation was higher, ranging from 89 to 100%. The results of this typing were compared with the clinical origin of the strains. Only strains of serogroups A, C, and D were isolated in 153 cases of antibiotic-associated diarrhea. This series included strains from three outbreaks; all the strains in two of the outbreaks belonged to serogroup C, and in the third, all the strains belonged to serogroup A. Strains of serogroups B, F, and G were only found in the stools of asymptomatic neonates or young children. In the latter samples, strains of serogroups A and D were found in the same ratio as in adults with antibiotic-associated diarrhea, but strains of serogroup C were seldom isolated. In patients treated with antineoplastic drugs and suffering from diarrhea, the distribution of the strains was the same as in cases of antibiotic-associated diarrhea.

Interaction between Clostridium difficile and polymorphonuclear leucocytes from the elderly and post-operative cancer patients: phagocytosis and bactericidal function

The ability of polymorphonuclear leucocytes (PMNs) from elderly healthy subjects and post-operative cancer patients to ingest and kill Clostridium difficile was studied. The percentage of phagocytosis in clinically healthy subjects aged 69-82 years ranged from 20.2 +/- 3.2 (mean +/- SD) to 34.1 +/- 4.2, depending on the strain of C. difficile, from 6.7 +/- 2.2 to 11.2 +/- 2.2 in post-operative cancer patients aged 65-69, and from 68.4 +/- 3.1 to 81.1 +/- 6.3 in young healthy control subjects. In both study groups, the defect was in part serum-associated. Although the total number of bacteria killed by the PMNs was reduced in the two study groups, the percentage of ingested bacteria killed by the PMNs was similar to that in the young healthy subjects. These differences were not noted when Escherichia coli and Staphylococcus aureus were used as the test organisms. The defective PMN function against C. difficile in the elderly and post-operative cancer patients described in this study may offer an explanation as to why these individuals are at high risk of developing antibiotic-associated colitis.

Clostridium difficile colitis in surgical patients

The clinical course of 75 patients with diarrhea and positive C. difficile toxin stool assays has been examined. The mean age of the patients was 68 years. Five of 25 surgical nursing units accounted for two thirds of the cases. Many patients were immuno-suppressed with cancer, sepsis, or diabetes mellitus. The median onset of diarrhea was 2.7 days after initial administration of antibiotics. Fever and leukocytosis were frequently seen. Diarrhea ceased in 30 percent of the patients after withdrawal of the offending antibiotics. The remainder required specific therapy with vancomycin, bacitracin, or metronidazole. Two deaths were directly attributable to C. difficile colitis. The hospital stay was prolonged in many patients. C. difficile colitis should be suspected in any patient in whom diarrhea develops during or after a course of antibiotics. Enteric precautions may prevent clustering in these cases and colonization in other susceptible patients.

Comparison of methods for recovery of Clostridium difficile from an environmental surface

Survival of Clostridium difficile in an aerobic environment is possible because of spore formation. When sodium taurocholate is substituted for the egg yolk of a selective medium, cycloserine-cefoxitin-fructose-agar (CCFA), enhanced recovery of C. difficile spores is shown. This selective medium (TCCFA) does not improve recovery of vegetative forms. In this study, dry and saline-moistened swabs, adhesive paddles, and Rodac plates containing CCFA and TCCFA were compared in their ability to recover C. difficile spores from an inoculated surface. Rodac plates grew 20 to 25 times as many spores on TCCFA as on CCFA. Saline-moistened swabs recovered fewer organisms than Rodac plates. Dry swabs and adhesive paddles rarely recovered spores. Prereduction of agar in an anaerobic chamber was not necessary for optimal spore recovery. Optimal growth of vegetative C. difficile required prereduced media. Agar prereduced for 2 h supported the growth of 12 C. difficile isolates as well as agar prereduced for 18 h. Vegetative cells of C. difficile survived for only 15 min in room air. Use of Rodac plates containing TCCFA is preferred for detection of C. difficile spores in the hospital environment.

Nosocomial spread of Clostridium difficile

Environmental transmission of Clostridium difficile, the causative agent of antibiotic-associated pseudomembranous colitis (PMC), has been supported by animal studies and implicated in spread of C. difficile among leukemic children receiving non-absorbable antibiotics. We report antibiotic-associated C. difficile-related colitis in two adults who shared a commode chair during hospitalization.

Trimethoprim/sulfamethoxazole-associated pseudomembranous colitis

A case of TMP/SMX-associated pseudomembranous colitis is described in a patient being treated for a urinary tract infection. Pseudomembranes are visualized on proctosigmoidoscopy, and stool cultures identified Clostridium difficile as the causative organism. The patient was treated successfully with oral vancomycin. A review of this infrequently reported adverse reaction to TMP/SMX is presented, emphasizing etiology and treatment.

Treatment of Clostridium difficile colitis and diarrhea with vancomycin

Toxigenic Clostridium difficle is the major cause of antibiotic-associated colitis and is susceptible to vancomycin at fecal concentrations achieved with oral therapy. The effect of oral vancomycin was studied in 16 patients with C. difficile-related diarrhea or colitis, 12 of whom had colitis documented by endoscopy, biopsy, and/or barium enema. Four patients had antibiotic-associated diarrhea and possibly antibiotic-associated colitis, because sigmoidoscopy either showed normal results (two patients) or was not performed (two patients). Nineteen episodes of diarrhea were treated with oral vancomycin in two dosage regimens for three to 14 days. Twelve patients received 2 g daily, and four patients initially received 1 g or less per day. Within 48 hours of the start of vancomycin therapy, 14 of 16 patients (87 percent) showed a decrease in temperature, abdominal pain and diarrhea. Diarrhea ceased completely within two days of the start of vancomycin in nine episodes, within three to seven days in six episodes, and within eight to 14 days in the remaining four episodes, and within eight to 14 days in the remaining four episodes. Diarrhea recurred in two of these patients (12 percent) when the drug inciting the initial episode of colitis was given again 42 days or more after vancomycin therapy was stopped; both patients responded again to retreatment with vancomycin. Oral vancomycin is an effective treatment of C. difficile-related colitis and diarrhea.