Web Annex Table 1. Description of study population for each centre
Study centre / Source population / Source of cases (primary and secondary) / Source of controls /Australia / People with a right to vote who reside in the study regions at the time of diagnosis as a case or of selection as a control, and capable of participating in a face-to-face interview in English. / Pathology laboratories servicing all neurosurgical units (for both brain tumours and acoustic neurinomas), surgeons specialising in treatment of acoustic neurinomas (for acoustic neurinomas), population-based cancer registries for parotid gland cancers and as back-up for malignant brain tumours (both Sydney and Melbourne) and for benign brain tumours and acoustic neurinomas (Melbourne only). / Controls were randomly selected from the electoral rolls covering the statistical divisions of Melbourne and Sydney. Only subjects actually resident in those divisions at the time of selection were included.
Canada:
Montreal / Citizens who reside in the study region at the time of diagnosis as a case or of selection as a control / Pathology and radiology departments, as well as medical archives in the following hospitals: McGill University Hospital Center, Centre Hospitalier de l'Université de Montréal, Jewish General Hospital, Hôpital du Sacré-Coeur de Montréal, Hôpital Charles Lemoyne, Cité de la Santé de Laval, St-Mary’s Hospital, Hôpital Maisonneuve-Rosemont, Montreal Neurological Institute / Electoral lists. In Quebec, the electoral lists are continually updated, from multiple sources of information available to the provincial government, such as drivers' permits and medical insurance cards.
Canada:
Ottawa / Residents of the study region at the time of diagnosis as a case or of selection as a control / Departments of neurosurgery, neuroradiology, neuropathology, and medical archives in the Ottawa Hospital, where treatment and diagnosis for the region is centralised (brain). Acoustic neurinoma and parotid gland cases were recruited through all ENT specialists in the Ottawa area. / Random digit dialling
Canada:
Vancouver / Residents of the study region at the time of diagnosis as a case or of selection as a control / Pathology, surgery, radiology, cytology and referral reports were made available to the study from the provincial BC Cancer Agency (three of the regional treatment centres), as well as reports from all hospitals located in the specified health and geographic regions that provide diagnostic, radiological and neurological services / The population-based BC Ministry of Health Client Registry which is a comprehensive database including all BC residents of three months or longer. Only subjects with a listed phone number were considered.
Denmark / Residents of the study region who speak Danish and have no previous history of cancer (excluding benign skin tumours) / Brain: 5 neurosurgical departments to which all brain tumour scans are referred - Rigshospitales, Odense, Armus, Alborg, Veole, Herlev.
Parotid and acoustic neurinoma: 7 Oncology departments that treat malignant parotid gland tumours and acoustic tumours. The DAHANCA group which unites the specialists working in these departments diagnoses and treats all parotid tumours, notified all cases to the study group.All acoustic tumours are referred to Dr Jens Thomsen, an ORL specialist, who is required to see all newly diagnosed patients within the year of diagnosis / Central Population Registry (all residents are registered)
Finland / Residents of the study region / Neuro-surgery departments of the university hospitals - Helsinki University Hospital, Tampere University Hospital, Turku University Hospital, Kuopio University Hospital, Oulu University Hospital. / Population Register Centre (all residents are registered)
France / Citizens who reside in the study regions at the time of diagnosis as a case or of selection as a control / Neurosurgery, neurology, otorhinolaryngology, neuroradiology departments in the following hospitals: Lyon (Hospices Civils, Centre Léon Bérard, Tonkin Clinic, Desgenettes Military hospital); Paris (Beaujon Hospital, La Pitié Hospital, Foch Hospital, Rottschild Foundation, Val de Grâce military hospital, Lariboisière Hospital, Ste Anne Hospital, Poissy-St Germain en Laye Hospital, Sud-Francilien Hospital, Henri Mondor Hôpital, Meaux Hospital, Marc Jacquet Hospital) 1 radiology service (Marseille) which treats patients from Lyon or Paris, the Departments of Medical Information (DIM). / A small series of controls were obtained using random digit dialling, but this proved to be a sub-optimal strategy. Most of the controls were thereafter selected from the electoral roles.
Germany / Residents of the three study regions with sufficient knowledge of the German language to undertake the interview / Bielefeld (Bielefeld Gilead-Hospital), Heidelberg (University of Heidelberg Hospital, University of Mannheim Hospital) and Mainz (University of Mainz Hospital); these hospitals host the only neurosurgical clinics of the study region. All neuropathology services of the study region were also used for case ascertainment. In addition, the ear, nose and throat clinic in Mainz was included for ascertainment of acoustic neurinoma. / Regional population registries (all residents are registered)
Israel / Jewish citizens of Israel. Controls were also matched on country of birth. / Brain and acoustic nerve tumor cases: all general hospital neurosurgical departments (7), and the two private hospitals in which neurosurgery procedures are performed. An additional primary source for ascertainment of acoustic nerve tumor cases was one expert who is the only radiosurgery neurosurgeon for acoustic neurinoma in Israel.
Parotid tumor cases: Archive records from general hospitals (22) and private hospitals (2) as well as pathology records from general hospitals (21)
To verify completeness of ascertainment medical and pathological records from all relevant institutions were screened periodically / National Population Registry (all residents are registered)
Italy / Residents of Rome at the time of diagnosis as a case or of selection as a control. / Sixty-six departments from 33 public or private hospitals in Rome, including 14 neurosurgical departments, 2 neurotraumatology department, 9 ear nose and throat surgery departments, 2 maxillo-facial surgery department, 9 general surgery department, 25 pathology laboratories, and 2 radiotherapy units.. / Population Registry of the Municipality of Rome (all residents are registered).
Japan / Residents of the study region at the time of diagnosis as a case or of selection as a control. / Department of neurosurgery of the participating health institutions: 23 of the 30 large hospitals in the Tokyo area which are responsible for treatment of 90% of the diseases of interest. / Controls are selected from the general population using a random digit dialling method, in which phone numbers for fixed home phones are generated randomly
New Zealand / Residents of the study regions for at least 6 months / 3 central neuro-surgical units - Auckland (Auckland Hospital, Mater Hospital), Christchurch (Christchurch Hospital), and Wellington (Wellington Hospital). / Electoral rolls for the municipalities included.
Norway / Residents of the defined regions for the participating hospitals (Oslo National Hospital, Oslo Regional, Bergen and Trondheim). Until recently there were clearly defined catchment areas for all hospitals and all people were obliged to go to a designated hospital. / Neurosurgical departments, discharge registries and outpatient clinics in the four participating hospitals for brain and parotid gland tumours. The hospital in Bergen specialises in acoustic neurinomas. The other participating hospitals are: Haukeland Sykehus, Regionsykehuset I Trondheim, Rikshospitalet and Ulleval Sykehus. / Population registry (all residents are registered)
Sweden / Residents of the study regions, able to understand Swedish, and not completely deaf prior to diagnosis or reference date / The neurosurgery, oncology, neurology, and otorhinolaryngology clinics at all hospitals in the study areas where these patients are treated, including Karolinska sjukhuset, Sahlgrenska sjukhuset, Universitetssjukhuset i Lund, and Norrlands Universitetsjukhus. Benign parotid gland tumours (pleomorphic adenomas) were also identified at Huddinge sjukhus, Sofiahemmet, Danderyds sjukhus and Sahlgrenska sjukhuset. Cancer registries covering the study areas were searched regularly. / Population registry (all residents are registered)
UK-North / Residents of the study regions / Central Scotland (Glasgow Royal Infirmary, Southern General Hospital – Glasgow, Western General Hospital – Edinburgh), Trent (Queen’s Medical Centre – Nottingham, Royal Hallamshire Hospital – Sheffield), West Midlands (Birmingham Heartlands, Midland Centre for Neurosurgery/ Neurology, Solihull Hospital, Kidderminster General Hospital, North Staffs Hospital Centre, New Cross Hospital – Wolverhampton, Corbett Hospital, Queen Elisabeth Hospital -Birmingham, Selly Oak Hospital – Birmingham, Royal Shrewsbury, Walsgrave) and West Yorkshire (Airedale Hospital, Leeds General Infirmary, Leeds St James, Pinderfield & Pontefract, Huddersfield Royal Infirmary and Halifax Royal Infirmary & General) / Random selection of a set of eligible controls from the entire general practice patient list either where the case was registered or a practice close by.
UK-South / Residents of the study region / Neurosurgical, oncological and neuropathology centres in hospitals within the Thames region, including the Royal Marsden Hospital. Charing Cross Hospital, Guy’s and St. Thomas Hospital, King’s College Hospital, Princess Royal Hospital, Royal Free Hospital, St. George’s Hospital, The National Hospital for Neurology and Neurosurgery, Oldchurch Hospital, and Mount Vernon Hospital, and small numbers of patients from many other hospitals. / General practice patient lists. An additional attempt has been made to recruit controls from the electoral rolls.
Web Annex Table 2. List of ICD codes for eligible diagnoses
/ ICD-9 / ICD-10 / ICD-O2, ICD-O3 / ICD-O2, ICD-O3 /
/ Benign or suspected malignant / Malignant /
Glioma
High grade / 191.0-191.9 / D33.0,
D43.0 - D43.9,
C71.0 - C71.9 / C71.0 - C71.9 / 9442/1* / 9380/3, 9381/3, 9382/3, 9390/3**, 9392/3, 9401/3, 9423/3+, 9440/3, 9441/3, 9442/3, 9451/3, 9460/3***, 9505/3
Low grade / 225.0, 237.5,
191.0-191.9 / D33.0,
D43.0 - D43.9,
C71.0 - C71.9 / C71.0 - C71.9 / 9383/1, 9384/1, 9394/1, 9421/1+++, 9444/1, 9505/1 / 9391/3, 9393/3++, 9400/3, 9410/3, 9411/3, 9420/3, 9424/3**, 9450/3
Unassigned / 191.0-191.9 / C71.0 - C71.9 / C71.0 - C71.9 / 9390/0, 9390/1* / 9430/3, 9480/3
Meningioma / 225.2, 237.6,
192.1 / D32.0, D32.9,
D42.0, D42.9,
C70.0, C70.9 / C70.0, C70.9 / 9530/0, 9530/1, 9531/0, 9532/0, 9533/0, 9534/0, 9535/0, 9537/0, 9538/1, 9539/1* / 9530/3, 9538/3, 9539/3
Acoustic neurinoma / 225.1 / D33.3 / C72.4 / 9560/0
Parotid gland tumour / 210.2, 235.0,
142.0, 142.9 / D11.0, D11.9,
D37.0, C07.9 / C07.9 / 8010/0, 8010/2, 8147/0, 8149/0**, 8190/0, 8200/0, 8211/0, 8290/0, 8310/0, 8410/0, 8430/1**, 8450/0, 8503/0, 8550/0, 8550/1, 8561/0, 8940/0, 8982/0 / 8010/3, 8020/3, 8021/3, 8041/3, 8050/3, 8070/3, 8082/3, 8140/3, 8147/3**, 8190/3, 8200/3, 8290/3, 8310/3, 8410/3, 8430/3, 8440/3, 8450/3, 8480/3, 8500/3, 8550/3, 8560/3, 8562/3**, 8940/3, 8941/3, 8980/3, 8982/3
* ICD-O rev 3 only
** not in ICD-9
*** not in ICD-O rev 3
+ 9422/3 and 9443/3 in ICD 9, 10 and ICD-O2 were merged with 9423/3 in ICD-O rev 3
++ 9393/3 was 9393/1 in ICD 9, 10 and ICD-O2
+++ 9421/1 was 9421/3 in ICD 9, 10 and ICD-O2
Web Annex Table 3. Description of study approach to participants
Australia / Study information specifically referred to mobile phones – AUSPHONE. / A member of the neurosurgery unit or the Cancer Registry initially contacted cases for consent. All subjects were sent an approach letter and information sheet, followed by phone call for an appointment. / All subjects were sent an approach letter and information sheet, followed by phone call for an appointment. If no telephone number was available, an interviewer visited the house.
Canada:
Montreal / No direct mention of mobile phones, but the introductory letter referred to environmental factors including occupation and, in two hospitals, use of electronic devices. / Once authorisation was obtained from the treating physician to contact the case, an introductory letter presenting the goal of the study was mailed, accompanied by consent form and a short questionnaire. An interviewer then contacted them for an interview. / An introductory letter presenting the goal of the study was mailed to subjects, accompanied by consent form and a short questionnaire. An interviewer then contacted them for an interview.
Canada:
Ottawa / The introductory letter mentions mobile phone use as one of a list of factors to be evaluated. / Cases identified while in hospital were approached by the attending physician or a nurse assigned to the study for the purpose and informed of the study. A research assistant or interviewer then contacted them for an interview. Back ascertained cases were contacted directly by the study centre. / Controls were selected from an Ottawa telephone database using random digit dialling methods. Potential controls were contacted by the consultant, Opinion Search, who enrolled participants in the Ottawa Hospital catchment area by asking participants to agree to be contacted by the study research coordinator and/or an interviewer.
Canada:
Vancouver / The introductory letter describes the purpose as assessing cancer risk from use of mobile phones. / All subjects were sent a letter of invitation to participate, with an information sheet about the Cancer Registry, its purpose and how we are able to use it to identify individuals as potential study participants. A research assistant or interviewer then contacted them for an interview. / In addition to information about the study, he letter to controls contained additional information explaining how their names had been obtained. A research assistant or interviewer then contacted them for an interview.
Denmark / The study was presented as a study of brain and acoustic nerve tumours in relation to mobile phone use / The nurse or the consultant presented the study to the patients. When a patient was referred to a department nurses were assigned to await the patient and obtain the interview. / The introductory letter, consent form and the leaflet were sent to the subjects by mail. Once the signed consent was received the subject was assigned to an interviewer.
Finland / The study was described as a case-control study of brain tumour aetiology, covering several possible factors such as mobile phones. / A study nurse approached the subject, obtained consent and arranged the interview. / First, a letter was mailed requesting the subjects to contact the local study nurse. Initially we used a mailed reminder, but due to low response rate we switched to follow-up contact by phone, whenever possible
France / The letter stated that “the aim of the study is to study the relations between health and environment”. / The treating physician presented the study to the cases. An introductory letter explaining the study, accompanied by a consent form was either given directly or mailed to the subject. On receipt of consent form the interviewer then contacted them for an interview. / An introductory letter explaining the study, accompanied by a consent form was mailed to the subject (two successive mailings if needed). On receipt of consent form the interviewer then contacted them for an interview.