follow immunocompromised contact group of a case of open ... · screeningprotocolforpatients...

Thorax 1995;50:863-868

Follow up of an immunocompromised contactgroup of a case of open pulmonary tuberculosison a renal unit

F A Drobniewski, J Ferguson, K Barritt, R M Higgins, M Higgon, D Neave,A H C Uttley, D O'Sullivan, A Hay

Public HealthLaboratory andDepartment ofMicrobiologyF A DrobniewskiA H C UttleyA Hay

Department of RenalMedicineK BarrittR M Higgins

Department ofRespiratory MedicineM Higgon

Department ofOccupational HealthD Neave

Dulwich and King'sCollege Hospitals,London, UK

Directorate of HealthPolicy and PublicHealth, Lambeth,Southwark andLewisham HealthCommission,London, UKJ FergusonD O'Sullivan

Reprint requests to:Dr F Drobniewski, DulwichPublic Health Laboratory,Dulwich Hospital, EastDulwich Grove, LondonSE22 8QF, UK.

Received 30 January 1994Returned to authors4 April 1994Revised version received18 April 1995Accepted for publication24 April 1995

AbstractBackground - The organisation, manage-

ment, outcome and cost of follow upof a large group of mainly immuno-compromised patients and healthcareworkers who were exposed to a staffmem-ber of a London renal unit with smear

positive pulmonary tuberculosis are de-scribed.Methods - Following British Thoracic So-ciety (BTS) guidelines, 576 close contactswere identified and divided into threegroups: (1) 303 renal patients including61 with renal transplants; (2) 90 surgicalpatients; and (3) 183 staff members.Screened contacts were interviewed, com-pleted a symptoms questionnaire, andwere offered a chest radiograph and Heafor Mantoux test if appropriate with re-

ferral to a chest physician if required.Results - Overall, 524 (85%) living contactshave been screened: 243 (97%) renal (firstscreening), 63 (70%) surgical, and 135(74%) staff contacts. Thirty one transplantpatients were prescribed isoniazid chemo-prophylaxis. Fifty two renal patients haddied before screening and 11 deaths oc-

curred after first interview. One case oftuberculosis epidemiologically related tothe index case was diagnosed on clinicalcriteria. A review of the case records and/or death certificates and entries on totuberculosis registers indicated no furthercases. The cost of the investigation was

estimated to be approximately £25 000, or

£44 per contact screened, with staff costscomprising 79% of the total.Conclusions - Undiagnosed tuberculosisin healthcare workers working with im-munosuppressed patients can lead to largeand expensive follow up studies. The ap-plicability ofthe 1990 and 1994 BTS guide-lines to the investigation of tuberculosisin an inunmunocompromised nosocomialgroup, and the role ofthe infection controldoctor and the consultant in Com-municable Disease Control in overlappingnosocomial and community incidents, are

discussed.(Thorax 1995;50:863-868)

Keywords: tuberculosis, renal, immunocompromised.

Mycobacterial diseases are a major health prob-lem in the world today. Tuberculosis due toMycobacterium tuberculosis is the predominant

infectious cause of morbidity and mortalityworldwide, infecting eight million and killingthree million people annually.'2 The decline inthe number of cases of clinical tuberculosis inthe developed and parts ofthe developing worldceased or reversed in the mid 1 980s andthe reasons for this are discussed in detailelsewhere. '2

Tuberculosis in immunosuppressed patients,including those undergoing renal trans-plantation, is associated with atypical clinicaland radiographic presentation, rapid diseaseprogression, and a high mortality. Tuberculosismay be relatively common in renal patients -

for example, it has been diagnosed in up to28% of renal patients undergoing maintenancedialysis in one centre.3 In two recent studiesand numerous case reports incidences oftuber-culosis of 1-7-3-5% in renal transplant centresin the United Kingdom and in Saudi Arabiaare reported; geographical and ethnic variationare reflected in the incidence of 1 1 5% reportedin India.4 These studies confirm the sus-ceptibility ofsome renal patients to tuberculosiscompared with the general population, thatdisseminated tuberculosis is more commonafter renal transplantation, that it frequentlyrequires invasive procedures for diagnosis, andoften leads to loss of the graft when im-munosuppressive therapy is discontinued.45

MethodsIn September 1993 a staffmember working parttime in the Renal Unit and Regional TransplantCentre of a London hospital was diagnosedwith smear positive cavitating pulmonarytuberculosis. Relevant past medical history ofthe index case included BCG vaccination in1974 and a domestic tuberculosis contact in1988 which had been followed up ap-propriately; Heaf testing at that time was neg-ative and BCG vaccine was administered. Asubsequent pre-employment health screen in1990 was unremarkable. No Heaf test wasperformed. The index case, a smoker, had hadpulmonary symptoms for most ofthe precedingsix months and was subsequently referred toa chest physician who diagnosed tuberculosisclinically, confirmed by cavitating lesions onchest radiography. Sputum was smear positiveand yielded a fully susceptible strain of Mtuberculosis. The staff member was successfullytreated with conventional combination anti-tuberculous chemotherapy. 1 January 1993 was

863

on March 11, 2021 by guest. P

rotected by copyright.http://thorax.bm

j.com/

Thorax: first published as 10.1136/thx.50.8.863 on 1 A

ugust 1995. Dow

nloaded from

http://thorax.bmj.com/

Drobniewski, Ferguson, Bamtt, Higgins, Higgon, Neave, Uttley, O'Sullivan, Hay

designated as the onset of potential infectiousexposure because of the long duration of symp-toms including productive cough and the ad-vanced changes on chest radiography.

ACTION AND ORGANISATIONAn Emergency Control ofInfection Committee(EICC) chaired by the hospital infectioncontrol doctor was established which includedthe local consultant in communicable diseasecontrol and representatives from the renal,respiratory, public health, microbiological,radiological, and occupational health depart-ments, together with representatives from themanagement of the Trust. Its role was to definethe protocol to be followed (through a smallerClinical Protocol Group), establish the extentand end point of the investigation, disseminateinformation to all relevant parties including themedia, and to undertake major financial andmanagerial decisions. Designated additionaltelephone lines for patient advice supported bymedical advisors and information sheets on

tuberculosis were installed. Patients were in-terviewed using a standardised questionnaireeither by a renal nurse dedicated to the contactgroup, a chest unit liaison sister, a clinicalmicrobiology registrar, an infection controlnurse or an occupational health advisor as ap-

propriate. Screening chest radiographs were

reviewed initially by staff in the departments ofrenal or chest medicine or clinical microbiology,but were all subsequently evaluated by a con-

sultant radiologist or chest physician. Allpatients or staff with suspicious symptoms or

chest radiographs were referred to the chestclinic for further evaluation.A small Data Management Group (DMG)

was established comprising a registrar in clinicalmicrobiology, senior registrar in public health,

Patients:* All contacted by letter, explanation, invitation for screening* At first visit:

Symptom questionnaire, BCG status/scar and chest radiographyHeaf test for immunocompetent patients onlyChemoprophylaxis with isoniazid if necessaryAll chest radiographs reviewed by chest consultantReview by chest clinic if necessary

* At follow up:Renal o Questionnaire

Chest radiographySurgical Reassurance

GP letter

Chest clinic if required

Staff:* Fact sheet and invitation for screening* Occupational health consultation:

Symptoms questionnaireHeaf testingChest radiographyChest clinic referral if necessary

* Former staff: local consultants in communicable disease control notifiedof contact

Screening protocol for patients and staff exposed to pulmonary tuberculosis.

chest liaison sister, renal nurse, infection con-trol nurse, occupational health advisor and in-formation officer. Its role was to co-ordinatedata collection, organise follow up of the sur-gical patient group, identify possible tuber-culosis cases, and exclude tuberculosis as acause of death in those who died beforescreening began. Case records of deceased in-dividuals, death certificates, laboratory reportsand tuberculosis notifications were evaluatedto exclude tuberculosis as a cause of death inthe study population. The consultants in com-municable disease control of areas in whichformer Trust staff members were now workingwere contacted to follow up staffusing the BTSguidelines.6 Repeat letters were sent to patientand staff non-attenders and subsequently totheir GPs. In all cases GPs were informedof their patient's exposure, the result of thescreening or lack of attendance, and helplinenumbers and advice were offered.The cost of screening was derived from in-

dividual departmental costs which were as-sessed by appropriate members of the EICCand the costs of meetings which were costedon the basis of the number of hours spent byindividuals using mid point salary scales of theWhitley Council for 1993-4 including 14% oncosts.

CONTACTSContacts were defined using the BTS guide-lines6 and formed four groups: (1) family con-tacts; (2) renal patients; (3) surgical patients;and (4) staff. All family contacts were followedup locally in accordance with BTS guidelines.No cases of clinical disease were recorded buttwo children were placed on six months iso-niazid chemoprophylaxis. This study is con-cerned with the last three groups.The follow up period was initially set at two

years6 (later modified), with the first follow update at three or six months since last contactand then at six or 12 months, 18 months, and24 months. The lengthy follow up period wasfelt necessary as diagnosis was likely to be moredifficult in immunocompromised renal patientsin whom the average age was 54 years, tuber-culin testing was unreliable, and in whom therewere frequently non-specific pulmonary symp-toms with abnormal chest radiographs. Followup was as described in the figure. Heaf testswere not offered to the renal contacts due tothe high false negative rate in this older andimmunosuppressed group.

Chemoprophylaxis with isoniazid 200 mg(plus 10 mg pyridoxine) was given for sixmonths to those identified as being at specialrisk - namely, patients transplanted or receivingimmunosuppression for rejection, patients whowere HIV positive, and those receiving cyclo-phosphamide as induction therapy for renalvasculitides.

Surgical contacts were followed up as in-dicated in the figure but they were reviewed intwo additional clinics, Heaf tested as ap-propriate, and offered chest radiography. Fol-lowing the initial screen and the lesser risk tothese contacts, surgical patients were asked to

864



j.com/


ugust 1995. Dow

nloaded from


Follow up of an immunocompromised contact group of a TB patient

Table 1 Renal contacts of index case screened in accordance with the protocol outlined inthe figure

First screening Second screening

Requiring screening 303Screened in district of residence 23Died before screening 52 11 (5 6%)Requiring screening locally 228 217Screened in study 220 196Signed disclaimer 5 (2 2%)' 4 (1-8%)Did not attend 3 (1-3%) 7 (3 2%)% screened 97% 95%

'Figures in brackets represent percentages of contacts requiring screening locally (n = 228 at firstscreening, n=217 at second screening).

Table 2 Symptoms, BCG history, and screening undertaken among different contactgroups (n = number completing all or part of the screening process)

Renal Surgical Staffcontacts contacts contacts

No. of questionnaires completed 220 57 135Mean age (years) 54 58Age range 17-85 18-91Male 122 (56%) 23 (40%) 36 (27%)Female 95 (44%) 34 (60%) 99 (73%)Symptoms:Cough 92 (42%) 24 (42%) 35 (26%)Phlegm 55 (25%) 17 (30%) 24 (18%)Bloody phlegm 3 (1%) 1 (2%) 1 (1%)Weight loss 25 (11%) 12 (21%) 4 (3%)Fever 42 (19%) 6 (11%) 9 (7%)Sweating 47 (21%) 13 (23%) 23 (17%)

BCG history 66 (30%) 20 (35%) 107 (79%)Chest radiography 217 (99%)' 52 (92%)d 91 (67%)Referral to chest clinic 9 (4 1%)b 0 (0%)C 1 (0 7)

aOne patient refused chest radiography at the time of screening, two had chest radiographyperformed at another hospital.This group was under regular review by the renal department and decisions regarding chemo-

prophylaxis/treatment were made without formal referral but in collaboration with the chestclinic.cAlthough no formal referrals were made, this group was reviewed by a chest liaison sister andclinical microbiologist with immediate access to the chest clinic.d Of the five patients who did not have screening chest radiographs, one was an inpatient andanother collapsed in the screening clinic with acute pancreatitis and was admitted - both hadnormal radiographs; three refused radiography.'No ages were recorded from staff contacts to preserve confidentiality.

refer themselves to their GPs if they developedany suspicious symptoms. Written advice tothis effect was given to all patients and theirGPs and details entered in the surgical casenotes.

Staff members with minimal contact werereassured and the remainder followed up by theoccupational health department as described inthe figure. All 27 family contacts resided out-side the district and were followed up by theirlocal contact tracing services.The protocol was modified by the EICC

in the light of proposed changes to the BTSguidelines7 that contacts should be followed upfor a period of 12 rather than 24 months sothat patients and staff who remained symptomfree, with no evidence of tuberculosis on chestradiography at 12 months, would be dischargedfrom follow up with notification to their GPs.

ResultsOverall, 576 patient and staff contacts wereidentified and 85% (543) were reviewed. Threehundred and three renal contacts were iden-tified including 61 (27 7%) with renal trans-plants, 36 (1 64%) on haemodialysis, 35(1 5-9%) on ambulatory peritoneal dialysis, and76 (34 6%) with general nephropathology; 52patients had died before screening and 23 werescreened by their local referring hospital leaving228 patients for local screening (table 1). A

further 11 patients died during the screeningprocess. The mean age, age range, reportedBCG status, and proportion screened referredfor chest radiography and to the chest clinicare described in table 2. Although 135 (74%)staff contacts were screened there were sig-nificant differences in the groups who attended- for example, only nine (33 3%) medical staffattended compared with 80-7% of the otherstaff groups (nursing, domestic, portering, etc)(X 26-77, p<0-001).The range of symptoms reported among the

contacts appears in table 2. Of the renal con-tacts, 132 (60%) reported at least one symptomand 85 (39%) two or more. Three patientsreported apparent haemoptysis but did not havetuberculosis on further review. Amongst thesurgical patients 34 (60%) reported at least onesymptom; one individual reported haemoptysis(related to a postoperative problem) and notuberculosis was detected. Forty five staff(33%) reported at least one symptom withone staff member reporting haemoptysis. Thisemployee had ceased employment with theTrust and was followed up by her new em-ployers who concluded that the temporaryhaemoptysis was due to severe acute bronchitisin a heavy smoker and that there was no evid-ence of tuberculosis.Knowledge of BCG vaccination status was

variable but staff contacts were generally ac-curate when status was confirmed by the pres-ence of a BCG scar. Few Heaf test results wereavailable for analysis on patients as no testswere performed on renal patients and manysurgical patients were elderly (rendering thetest unreliable), or had concurrent illness, orrefused to have the test administered or read.Only one staff contact individual with a Heafgrade 3 or 4 had definitely not been vaccinated.Subsequent chest radiography and follow uphad not indicated tuberculosis. No other staffgave Heaf tests of grade 3 or 4. The staff whowere Heaf tested were given BCG if negativebut continued working if asymptomatic with aprior history of BCG and/or presence of a scar.The medical records of the 48 staff non-

attenders were reviewed indicating that 36(75%) had been BCG vaccinated previously.Twenty two (45 8%) had been Heaf testedduring their employment screening with 20producing a Heaf 1 or 2 reaction and twoproducing a Heaf grade 3 reaction. Subsequentchest radiography and follow up at the timehad not indicated tuberculosis. Screening hasbeen definitely carried out on at least 10 of the48 individuals (21%) who have left the Trust'semployment.

OUTCOMESixty three renal patient contacts died duringthe contact period and between screenings.None of the deaths was believed to be due totuberculosis following review of case notes,death certificates, or discussion with the at-tending medical staff or GP. One case of pul-monary tuberculosis was identified during thescreening process. A 55 year old white womanreceiving haemodialysis who had no symptoms

865



j.com/


ugust 1995. Dow

nloaded from


Drobniewski, Ferguson, Bamitt, Higgins, Higgon, Neave, Uttley, O'Sullivan, Hay

Table 3 Final costs associated with the investigation of staff and patient contactsa

Department or group Cost Staff costs as % of total

Emergency Infection Control Committee (EICC) £2407 100Clinical Protocol Group £167 100Data Management Group (DMG) £579 100Renal unit £6499 92-5Chest unit £1143 100Radiology department' £4445 87 1cOccupational health department £3161 94-8General management £693 60-8Public health and public health laboratory service £5715 1oodTotal £25388

'Direct costings are derived from mid point salary scales of the Whitley Council for 1993-4including 14% on costs. Where costs were estimated for meetings such as those of the EICConly the hourly rate for all those attending was assessed; no allowance was made for travellingtime or costs or cancelled clinical sessions.'Apart from meetings a single costing price was used to cover the radiograph, radiographer, andradiologist reporting time.'An individual package cost of radiograph, radiographer, and radiologist time was provided.dPhotocopying costs and computer facilities were provided from within the routine budget.

and a normal chest radiograph three monthsafter her contact with the index case in October1993 subsequently reported a productive coughwith fever on her second screening in January1994. She had no other history of contact withtuberculosis, had not been immunised withBCG, and was not on immunosuppressivedrugs. She was also hepatitis C positive. A chestradiograph now revealed apical shadowing inboth zones consistent with tuberculosis, butcultures of sputum, blood and bronchial wash-ings failed to grow M tuberculosis. A clinicaldiagnosis of tuberculosis was made, she com-menced antituberculous therapy, and has nowrecovered. In all, 31 immunocompromisedrenal patients received a course of isoniazidchemoprophylaxis, 18 (16 ofwhom were trans-plant patients) at the initial renal unit and 13transplant patients at a second centre sharingcare with this unit.

Tuberculosis was not detected in any of thesurgical patients, although several patients re-ported previous tuberculosis or past exposureand had radiological signs consistent with oldcalcified tuberculosis. Two pulmonary malig-nancies were identified incidentally and ap-propriate referrals made. No deaths fromtuberculosis were identified during the followup period.

COSTINGThe final costs associated with the investigationare given in table 3. The cost of meetings ofthe EICC and DMG was £3 153 but the formercost almost five times as much as the latter,reflecting both its larger size and the greaterseniority of its members. The overall cost ofthe study was approximately £25 000 or £44per contact screened. Total staff costs wereestimated to be 78% of the total costs.

DiscussionThis study raises several issues regarding theadvisability and practicalities of managing alarge contact group using the 1990 BTS guide-lines.6 No guidelines were available for trans-plant patients and these were formulated duringthe study. Many of these recommendationsappear in the revised 1994 BTS guidelines7 in

which a less proactive response is advised forall but the immunosuppressed.Many contacts were identified due to the

prolonged period ofexposure in a surgical wardand a busy renal unit carrying out ap-proximately 70 transplants per year with 350transplant patients on follow up, 65 haemo-dialysis patients, and 135 patients on con-tinuous ambulatory peritoneal dialysis underits care at any one time.The DMG organised and co-ordinated data

collection and reported to the EICC. In prac-tice, although it worked extremely efficiently, itwas called upon to make management decisionsbeyond its original remit. Authority to makeday-to-day decisions vested in a small groupwould have expedited matters, particularly ininterpreting BTS guidelines in the context oflocal conditions.

CONFIDENTIALITY AND EXECUTIVERESPONSIBILITYInevitably the identity of the index case wasknown amongst the staff member's peer groupbut protection of the staff member's identitywas considered to be ofparamount importanceethically. The organisers of the investigationensured that confidentiality was preserved, butwith complete openness regarding the natureof the disease and the likely risks to patientsand staff.The issue of lead executive responsibility for

the outbreak was highlighted by this study.Although a good working relationship existedbetween the consultant in communicable dis-ease control and the infection control doctor,the exact roles in any similar incident needto be explicitly defined from the outset. Inprinciple, an infection control doctor deals withany outbreak involving the hospital alone, andthe consultant in communicable disease controlwould normally have executive responsibilityin any incident involving a notifiable disease inthe community or in a serious public healthproblem originating in a hospital but whichaffects the population at large.8

COSTS AND RESOURCESResources from central Trust funds were madeavailable originally for telephone lines andstaffing but costs for additional staff in the keydepartments involved in the handling of theincident came from their own budgets onceanticipated public and media interest did notmaterialise.The immunocompromised status ofmany of

the contacts increased costs as a broad rangeof screening tools had to be deployed. Theprevalence of reported symptoms from thequestionnaire if used alone, for example, wasof low predictive value for both patient groupsbut was of value in screening staff for tuber-culosis. Immunosuppression produces anergyto tuberculin and renal patients frequently hadatypical symptomatology and abnormal chestradiography reducing the individual discrim-inatory power ofMantoux testing, clinical, andradiological examination.

866



j.com/


ugust 1995. Dow

nloaded from


Follow up of an immunocompromised contact group of a TB patient

CONTACTSRenal contacts were easy to identify as thepatients were reviewed regularly and the renaldepartment keeps excellent computerised rec-ords which enabled lists to be created quickly.Complete staff lists were also obtained withoutdifficulty. The production of complete lists ofsurgical patients was surprisingly difficult des-pite the fact that many of the patients were

extra-contractual referrals requiring billing tothe originating health authority. It also becameclear that, unless a patient dies in hospital, thereis no simple mechanism to ascertain whichpatients have died and so prevent families re-

ceiving inappropriate requests for deceased rel-atives to attend for screening.The appointment of a dedicated staff nurse

in the renal unit with personal knowledge ofpatients to co-ordinate the contact tracing was a

key factor in the high success rate for screening.The principle ofa single individual with generalresponsibility for managing a contact group isa cost effective approach.

RecommendationsOne case of tuberculosis was discovered equat-ing to a cost of £44 per contact screened or

£25 000 per case detected. Two cases of lungcancer were diagnosed earlier than if screeninghad not occurred and staff members who hadnot received BCG were identified. The oc-

cupational health department has changed itspolicy to conform with the latest BTS guide-lines and closer liaison and regular policy revieware planned. The Trust has reinforced theimportance of attendance for screening in anyoutbreak situation.However, guidelines are promulgated in part

on the basis that most cases of clinical tuber-culosis are due to reactivation of earlier in-fection. Two recent reports have suggested that30-40% of infections recorded in New YorkCity and San Francisco during a two yearperiod were due to recent infection.910 Smallet al9 described one non-compliant individualwho, on the basis of restriction fragment lengthpolymorphism (RFLP) typing, infected 6% ofall identified cases in San Francisco despitethe efforts of an excellent tuberculosis controlprogramme which ensured that 95% ofpatientscompleted treatment during the study period.Nosocomial spread of M tuberculosis infectionin a renal transplant unit in the USA in whichfive patients died was recently proved usingRFLP analysis,"1 and the same technique wasused to disprove a nosocomial outbreak in aLondon renal unit.'213 Molecular techniquesare likely to be of increasing value in analysingnosocomial as well as community outbreaks.'4If these findings are generally applicable thencontact tracing with molecular analysis ofposit-ive cultures may become more common.

In a similar incident the 1994 BTS guide-

lines,7 which have superseded the 1990guidelines,6 should form the basis for its in-vestigation, supplemented by the followingrecommendations:

(1) clear executive responsibility for the in-

cident must be established from the out-set;

(2) adequate financial resources from centralTrust/hospital funds must be identifiedfrom the outset;

(3) a small task force must be establishedwith staff assigned full time as required.The group should have sufficient del-egated authority from the EICC to con-duct all actions needed to deal with theincident;

(4) it must be seen as a management prioritythat accurate lists of contacts should beprovided;

(5) all possible efforts must be made toidentify deceased patients to ascertainwhether death was related to tuberculosisand to prevent distress to families in-advertently receiving notifications forscreening;

(6) patients and GPs must be fully informedof the incident, the risk of infection andlikely treatment needed, and must haveready access to further information;

(7) patients and GPs must be informed ofthe conclusions of individual in-vestigations and when the follow upperiod is terminated;

(8) staff confidentiality must be preserved onethical grounds and to encourage staff tocome forward;

(9) monitoring, data collection, and analysismust be standardised;

(10) local guidelines should be formulated sothat an action plan is available that canbe instituted quickly;

(11) staff contacts must be aware that theyhave a responsibility to attend for screen-ing and senior line managers must maketime available regardless of other com-mitments for staffto attend appointmentswith occupational health or designatedmedical or nursing staff.

The authors thank the staff of the renal unit, particularly MsMonica McSharry and Professor B Hendry, Professor JMoxhamand Dr B Gray, and Ms Kathy Watkins, Ms Ann Doherty, MrJamil Choglay and Ms Linda Creegan for their help and supportin the conduct of this investigation.

1 Rodrigues LC, Smith PG. Tuberculosis in developing coun-tries and methods for its control. Trans R Soc Trop MedHyg 1990;84:739-44.

2 Bloom BR, Murray CJL. Tuberculosis: commentary on are-emergent killer. Science 1992;257: 1055-64.

3 Mitwali A. Tuberculosis in patients on maintenance dialysis.Am JKidney Dialysis 1991;18:579-82.

4 Higgins RM, Cahn AP, Porter D, Richardson AJ, MitchellRG, Hopkin JM, et al. Mycobacterial infections after renaltransplantation. QJ Med 1991;78:145-53.

5 Qunibi WY, al-Sibai MB, Taher S, Harder EJ, de Vol E,al-Furayh 0, et al. Mycobacterial infections after renaltransplantation - report of 14 cases and review of theliterature. QJYMed 1990;77:1039-60.

6 Joint Tuberculosis Committee of the British Thoracic So-ciety. Control and prevention of tuberculosis in the UnitedKingdom: Code of Practice 1994. Thorax 1994;49: 1193-200.

8 NHS Management Executive. Public health: responsibilitiesof the NHS and the roles of others. HSG (93) 56, 1993.

9 Small PM, Hopewell PC, Singh SP, Paz A, Parsonnet J,Ruston DC, et al. The epidemiology of tuberculosis inSan Francisco - a population-based study using con-ventional and molecular methods. N Engl J Med 1994;330:1703-9.

10 Alland D, Kalkut GE, Moss AR, McAdam RA, Hahn JA,Bosworth W, et al. Transmission of tuberculosis in NewYork City - an analysis by DNA fingerprinting and con-ventional epidemiological methods. N Engl Jf Med 1994;330:1710-6.

11 Jereb JA, Burwen DR, Dooley SW, Haas WH, Crawford

867



j.com/


ugust 1995. Dow

nloaded from


Drobniewski, Ferguson, Bamitt, Higgins, Higgon, Neave, Uttley, O'Sullivan, Hay

JT, Geiter LJ, et al. Nosocomial outbreak of tuberculosisin a renal transplant unit: application of a new techniquefor restriction fragment length polymorphism analysis ofMycobacterium tuberculosis isolates. Infect Dis 1993;168:1219-24.

12 Bendall RP, Drobniewski FA, Jayasena D, Nye PM, UttleyAHC, Scott GM. Restriction fragment length poly-morphism analysis rules out cross-infection among renalpatients with tuberculosis. Hosp Infect 1995 (in press).

13 Drobniewski FA, Bendall R, Nye P, Jayasena S, Scott G,Uttley A. Use ofrestriction fragment length polymorphism(RFLP) analysis of Mycobacterium tuberculosis to rule outcross-infection on a renal unit. Proceedings of the 3rd In-ternational Conference ofthe Hospital Infection Society, London1994, abstract F8/2.

14 Drobniewski FA, Kent RJ, Stoker NG, Uttley AHC.Molecular biology in the diagnosis and epidemiology oftuberculosis. Hosp Infect 1994;28:249-63.

868



j.com/


ugust 1995. Dow

nloaded from


follow immunocompromised contact group of a case of open ... · screeningprotocolforpatients...

Documents