Supplementary materials for:



Supplementary materials for: The utilization and safety of umeclidinium and umeclidinium/vilanterol in UK primary care: a retrospective cohort studyGema Requena, Daniel Dedman, Jennifer K. Quint, Rebecca E. Ghosh, Rachael Williams, Jeanne M. PimentaSources of dataPrimary care dataClinical Practice Research Datalink (CPRD GOLD) served as the main primary care database for development and benchmarking of analyses. It contains data extracted from Vision Patient Record Management software and includes anonymized longitudinal electronic health records (EHR) of patients registered with contributing primary care practices across the UK. The CPRD GOLD database covers approximately 3.6% of the current UK population, including practices in England, Northern Ireland, Scotland and Wales. As of August 2018, there were 740 general practices and 15.5 million acceptable (research quality) patients in CPRD GOLD, of which 4.3 million were currently registered at the start of the study accrual period on July 1, 2014. It is representative of the UK population with data collected from general practices since 1987.The Health Improvement Network (THIN) database contains the EHR of 11.1 million patients collected from 562 general practices in the UK, covering 6.2% of the UK population. All data are fully anonymized, processed and validated by CSD Medical Research UK. It is representative of the UK population with data collected since 1994. Some general practices contributed data to both the CPRD GOLD and THIN databases. All data from these practices were maintained in CPRD GOLD and excluded from the THIN data extract to avoid double counting.Linked Hospital Episodes Statistics Admitted Patient Care (referred to as HES)HES contains details of episodes of care delivered by (or on behalf of) the National Health Service (NHS), to patients admitted to hospitals in England. A wide range of information is recorded including clinical information about diagnoses (up to 20 per episode, coded using International Classification of Diseases, Tenth Revision [ICD-10]), and operations or procedures (up to 24 per episode, coded using OPCS Classification of Interventions and Procedures version 4 [OPCS-4]); patient information (age group, gender, and ethnicity); and administrative information, including dates and methods of admission and discharge. Data included in for this study covered admissions between April 1, 1997 and December 31, 2017.Linked death registrations from the Office of National Statistics (ONS)The linked death registrations include information derived from death certificates of all deaths registered in England, including date of death, and underlying and contributory causes of death (coded using ICD-10). Data included in this study covered deaths registered between January 2, 1998 and February 13, 2018.Linkage mechanismHES and ONS death registrations data were linked to CPRD GOLD primary care records of patients from 411 English general practices using a multistep deterministic algorithm, based on an exact match of NHS number plus at least one or more of gender, postcode of residence, date of birth (full or partial). The linkage was undertaken by NHS Digital which acted as a ‘trusted third party’ for CPRD and held the linkage identifiers for all three datasets.MethodsInclusion criteriaPatients were required to have ≥12 months of continuous data recording prior to index date. Patients were followed from their index date until censoring at death, leaving their general practitioner (GP) practice, the practice’s last collection date, or end of follow-up on June 30, 2017. Current exposure to index medication was defined as person-time from the index date until the earliest of the censoring date or discontinuation date. Discontinuation was defined as a break of ≥91 days between prescriptions or complete cessation in prescribing of the index medication. Among patients who discontinued index medication, all person-time from the discontinuation date up to the censoring date, or up to the date of resumption of the index medication (if applicable), was classified as not currently exposed. Drugs included in the ‘Other LABD’ groupsalmeterol, salmeterol+fluticasone, formoterol, formoterol+budesonide, formoterol+aclidinium, formoterol+beclometasone, formoterol+fluticasone, tiotropium, glycopyrronium, glycopyrronium+indacaterol, aclidinium, indacaterol, olodaterol, vilanterol+fluticasone, tiotropium+olodaterolMeasures of adherenceMedication possession ratio (MPR) estimates the amount of medication that was available to a patient during the first period of continuous use of the medication. MPR was only calculated for patients with ≥12 complete months of follow-up and at least 2 prescriptions for the index medication during the first period of continuous use. MPR = number of days’ supply of index medication between index and last prescription divided by the total number of days between index date and last prescription date issued before end of follow-up or first discontinuation date. The number of days’ supply was calculated by multiplying the number of prescriptions issued (not including the last prescription) by the assumed duration of 30 days. Adherence was defined as MPR ≥80%. Proportion of days covered (PDC) estimates the proportion of days in the year after index date in which the patient had index medication available. PDC was calculated for patients with ≥365 days of follow-up from the index date. PDC = number of days with index medication available (ie, covered) divided by 365 days. Number of days covered was calculated by multiplying the number of prescriptions in the period (including the last prescription) by 30 days. All prescriptions issued during the 365-day period were included; the total number of days covered was capped at 365. Patients with a single prescription for index medication were included in the PDC calculation. Data quality controlsThe standard operating procedures of CPRD were followed, including internal quality audits; secure storage and backup of confidential data and study documentation; quality control procedures for programming, and requirements for senior scientific review. All patients are required to have data of acceptable research quality according to CPRD standards. All statistical programming was independently reviewed by a second analyst.Supplementary Table 1. Patient characteristics at index in the linked cohortCharacteristicUMEC (n=547)UMEC/VI (n=512)Other LABD (n=10,590)Age at index, mean years (SD)70.5 (11.3)70.1 (11.3)65.7 (15.8)Age category, years, n (%)<18 years0─40─4242 (2.3)18?34 years0─40─4232 (2.2)35?64 years139 (25.4)134 (26.2)3680 (34.8)≥65 years 404 (73.9)376 (73.4)6436 (60.8)Female, n (%)245 (44.8)233 (45.5)5277 (49.8)Smoking status, n (%)Current smoker208 (38.0)190 (37.1)3807 (36.0)Ex-smoker282 (51.6)273 (53.3)4799 (45.3)Never smoker57 (10.4)48 (9.4)1854 (17.5)Missinga0 (0)1 (0.2)130 (1.2)BMI mean kg/m2 (SD)27.6 (6.4)28.0 (6.4)28.0 (6.7)Underweight (<18.5 kg/m2), n (%)26 (4.8)23 (4.5)432 (4.1)Normal (18.5?24.9 kg/m2), n (%)178 (32.5)154 (30.1)3110 (29.4)Overweight (25.0?29.9 kg/m2), n (%)185 (33.8)165 (32.2)3261 (30.8)Obese (≥30 kg/m2), n (%)151 (27.6)162 (31.6)3308 (31.2)Missinga, n (%)7 (1.3)8 (1.6)479 (4.5)Comorbiditiesb, n (%)Cardiovascular disease371 (67.8)316 (61.7)6352 (60.0)Prescription for a beta-blocker95 (17.4)90 (17.6)2004 (18.9)Pneumonia43 (7.9)36 (7.0)791 (7.5)Gastroesophageal disease119 (21.8)125 (24.4)2428 (22.9)Diabetes106 (19.4)82 (16)1975 (18.7)Kidney disease127 (23.2)103 (20.1)1802 (17.0)Cancer74 (13.5)84 (16.4)1405 (13.3)‘Moderate’ and ‘Severe’ COPD exacerbations in the previous yearc Rate per 1000 person-years (95% CI)1470 (1372, 1575)1111 (1024, 1206)1029 (1010, 1048)0 events, n (%)188 (40.4)222 (52.0)4873 (52.3)1 event, n (%)154 (27.8)144 (25.2)2937 (26.1)≥2 events, n (%)205 (31.8)146 (22.9)2780 (21.6)Dyspnea, mean (SD)2.8 (1.1)2.7 (0.9)2.6 (1.0)MRC grade 1, n (%)43 (9.9)33 (8.4)548 (9.8)MRC grade 2, n (%)149 (34.3)141 (35.8)2213 (39.4)MRC grade 3, n (%)138 (31.7)145 (36.8)1822 (32.4)MRC grade 4, n (%)79 (18.2)63 (16.0)874 (15.6)MRC grade 5, n (%)26 (6.0)12 (3.1)163 (2.9)Missinga, n (%)112 (20.5)118 (23.1)4970 (46.9)FEV1 percent predicted, mean (SD)56.1 (18.9)58.1 (20.1)60.9 (19.7)Mild, Grade 1 (≥80%), n (%)47 (10.7)43 (10.6)1068 (15.3)Moderate, Grade 2 (≥50% to <80%), n (%)222 (50.5)223 (55.1)3859 (55.2)Severe, Grade 3 (≥30% to <50%), n (%)139 (31.6)114 (28.2)1703 (24.4)Very severe, Grade 4 (<30%), n (%)32 (7.3)25 (6.2)356 (5.1)Missinga, n (%)107 (19.6)107 (20.9)3604 (34.0)FEV1/FVC ratio, mean (SD)57.3 (15.1)59.2 (15.5)62.8 (15.7)<70%, n (%)330 (84.8)309 (80.9)4331 (70.6)≥70%, n (%)59 (15.2)73 (19.1)1805 (29.4)Missinga, n (%)158 (28.9)130 (25.4)4454 (42.1)Ranges are shown where <5 events were reported; apercentages were calculated separately for those with missing and without missing data; bpatients who had ever had a diagnosis or who had a prescription for beta-blockers in the previous 12 months; caccording to primary care data and/or secondary care.BMI, body mass index; CI, confidence interval; COPD, chronic obstructive pulmonary disease; FEV1, forced expiratory volume in 1 second; FVC, forced vital capacity; LABD, long-acting bronchodilator; MRC, Medical Research Council; SD, standard deviation; UMEC, umeclidinium; UMEC/VI, umeclidinium/vilanterol.Supplementary Table 2. Patient characteristics by index treatment and diagnosis in the primary care cohortNew users of UMEC (N=3875)New users of UMEC/VI (N=2224)CharacteristicCOPD (n=3604)Asthma (n=130)Other (not COPD/asthma) (n=141)COPD (n=2029)Asthma (n=69)Other (not COPD/asthma) (n=126)Age at index date, mean years (SD)69.0 (10.3)60.2 (15.9)69.8 (13.8)69.3 (10.2)62.3 (15.0)68.4 (14.3)Female, n (%)1829 (50.8)75 (57.7)80 (56.7)947 (46.7)47 (68.1)61 (48.4)Smoking status, n (%)Current smoker1555 (43.2)42 (32.3)59 (42.1)819 (40.4)16 (23.2)48 (38.1)Ex-smoker1788 (49.6)44 (33.9)54 (38.6)1080 (53.3)26 (37.7)46 (36.5)Never smoker 261 (7.2)44 (33.9)27 (19.3)129 (6.4)27 (39.1)32 (25.4)Missinga0 (0)0 (0)1 (0.7)1 (0.1)0 (0)0 (0)BMI, mean kg/m2 (SD)27.8 (6.4)30.9 (7.9)27.6 (7.1)28.3 (6.3)30.3 (6.7)28.8 (6.8)‘Moderate’ COPD exacerbations, rate per 1000 person-years (95% CI)b1116 (1082, 1151)n/an/a789 (751, 829)n/an/a0 events, n (%)1687 (46.8)1136 (56.0) 1 event, n (%)904 (25.1)513 (25.3)2+ events, n (%)1013 (28.1)380 (18.7)Dyspnea (MRC Grade), mean (SD)2.8 (1.0)b2.71 (1.0)c3.4 (0.9)d2.7 (0.9)2.7 (0.8)2.6 (0.7)FEV1 percent predicted, mean (SD)57.9 (19.2)67.9 (18.8)64.1 (21.2)59.6 (18.7)76.6 (21.7)70.1 (22.2)Grade 1 (≥80%), n (%)312 (11.9)9 (24.3)8 (22.2)210 (13.4)3─610 (27.8)Grade 2 (≥50% to <80%), n (%)1387 (53.0)21 (56.8)19 (52.8)886 (56.3)6─920 (55.6)Grade 3 and 4 (<50%), n (%)917 (35.1)7 (18.9)9 (25.0)477 (30.3)1─46 (16.7)Missinga, n (%)988 (27.4)93 (71.5)105 (74.5)456 (22.5)49─53 (76.8)90 (71.4)FEV1/FVC ratio, mean (SD)58.6 (15.9)70.3 (11.8)68.6 (13.1)60.0 (14.0)73.7 (12.5)73.0 (15.0)<70%, n (%)1722 (79.9)16 (48.5)14 (50.0)1112 (79.1)1─414 (41.2)≥70%, n (%)433 (20.1)17 (51.5)14 (50.0)294 (20.9)9─1220 (58.8)Missinga, n (%)1449 (40.2)97 (74.6)113 (80.1)623 (30.7)56 (81.2)92 (73.0)Ranges are shown where <5 events were reported; apercentages were calculated separately for those with missing and without missing data; bn=2991; cn=14; dn=13.BMI, body mass index; COPD, chronic obstructive pulmonary disease; CI, confidence interval; FEV1, forced expiratory volume in 1 second; FVC, forced vital capacity; MRC, Medical Research Council; SD, standard deviation; UMEC, umeclidinium; UMEC/VI, umeclidinium/vilanterol. Supplementary Table 3. Cardiovascular, respiratory, and all-cause mortality outcomes in new users of UMEC and UMEC/VI during all available follow-up, including time while not currently exposed to index medication, for the primary care cohortUMEC (n=3875; 4962.6 person-years)UMEC/VI (n=2224; 3019.6 person-years)Outcome Total eventsa, nIR per 1000 patient-years (95% CI)b,cTotal eventsa, nIR per 1000 patient-years (95% CI)b,cMIAll patients366.5 (4.4, 9.1)237.3 (4.6, 11.1)With ICS use at index date236.0 (3.7, 9.3)1─4not calculatedNo ICS use at index date137.4 (3.8, 12.9)19─22not calculatedStroke All patients24532.0 (27.2, 37.5)14230.4 (24.5, 37.4)With ICS use at index date14427.0 (21.6, 33.2)1317.0 (7.3, 33.4)No ICS use at index date10142.4 (32.9, 53.9)12933.0 (26.2, 40.9)Newly diagnosed CHFdAll patients6013.1 (10.0, 16.8)3613.0 (9.1, 17.9)With ICS use at index date4113.4 (9.6, 18.2)1─4not calculatedNo ICS use at index date1912.4 (7.5, 19.4)32─35not calculatedPneumonia (primary care)All patients468.9 (6.5, 12.0)185.3 (3.0, 8.6)With ICS use at index date349.7 (6.6, 13.6)58.4 (2.3, 21.5)No ICS use at index date127.4 (3.8, 12.9)134.7 (2.4, 8.3)Moderate AECOPDcAll patients4663940 (899, 983)2241739 (692, 789)With ICS use at index date38481156 (1102, 1212)5671209 (1059, 1381)No ICS use at index date815503 (455, 557)1674655 (608, 705)Death (all causes)All patients27154.6 (48.3, 61.5)17658.3 (50.0, 67.6)With ICS use at index date18154.3 (46.7, 62.9)2450.2 (32.2, 74.7)No ICS use at index date9055.2 (44.4, 67.8)15259.8 (50.7, 70.1)aIncludes first and subsequent events. Ranges are shown where <5 events were reported; bincidence based on first event only; cfor AECOPD, all events were included; dpatients with prior history of CHF were excluded.AECOPD, acute exacerbations of COPD; CHF, congestive heart failure; CI, confidence interval; COPD, chronic obstructive pulmonary disease; ICS, inhaled corticosteroid; IR, incidence rate; MI, myocardial infarction; UMEC, umeclidinium; UMEC/VI, umeclidinium/vilanterol.Supplementary Table 4. On-treatment cardiovascular, respiratory, and all-cause mortality outcomes in new users of UMEC/VI and UMEC with and without concurrent treatment with other COPD maintenance therapy (primary care cohort)UMEC (n=3875; 3508.7 person-years)UMEC/VI (n=2224; 1768.3 person-years)Outcome Total eventsa, nIR per 1,000 patient-years (95% CI)b,cTotal eventsa, nIR per 1,000 patient-years (95% CI)b,cMIWith concurrent treatment196.4 (3.7, 10.2)1─4not calculatedNo concurrent treatment88.4 (3.4, 17.4)106.0 (2.7, 11.4)StrokeWith concurrent treatment12428.6 (22.5, 35.9)1027.2 (11.0, 56.1)No concurrent treatment4738.2 (25.9, 54.2)7431.0 (22.7, 41.4)Newly diagnosed CHFdWith concurrent treatment3614.6 (10.3, 20.3)1─4not calculatedNo concurrent treatment1215.3 (7.9, 26.7)1712.2 (7.1, 19.5)Pneumonia (primary care) With concurrent treatment227.9 (4.9, 12.0)1─4not calculatedNo concurrent treatment1─4not calculated1─4not calculatedModerate AECOPDcWith concurrent treatment30251143 (1085, 1203)3021113 (964, 1286)No concurrent treatment384509 (442, 585)973664 (605, 729)Death (all causes)With concurrent treatment7728.8 (22.7, 36.0)934.8 (15.9, 66.0)No concurrent treatment2833.6 (22.3, 48.5)5335.1 (26.3, 45.9)aIncludes first and subsequent events. Ranges are shown where <5 events were reported; bincidence based on first event only; cfor AECOPD, all events were included; dpatients with prior history of CHF were excluded.AECOPD, acute exacerbations of COPD; CHF, congestive heart failure; CI, confidence interval; COPD, chronic obstructive pulmonary disease; ICS, inhaled corticosteroid; IR, incidence rate; MI, myocardial infarction; UMEC, umeclidinium; UMEC/VI, umeclidinium/vilanterol.Supplementary Figure 1. Algorithm defining treatment patterns in patients who were not taking a concomitant COPD maintenance therapy at index dateCOPD, chronic obstructive pulmonary disease; UMEC, umeclidinium; UMEC/VI, umeclidinium/vilanterol. ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download