da Silva AG, Vieira JGH, Kunii IS, et al. The effects of discontinuing long term alendronate therapy in a clinical practice setting. Arquivos Brasileiros de Endocrinologia e Metabologia 2011;55:272-78
PubMed link: https://www.ncbi.nlm.nih.gov/pubmed/21779630
Full text article: http://www.scielo.br/scielo.php?script=sci_arttext&pid=S0004-27302011000400006
| Methods | Study design: Prospective cohort study
Number of groups: Three groups |
| Participants | Number of participants: 88 participants
· Intervention group 1: 40 enrolled, 35 completed · Intervention group 2: 25 enrolled, 19 completed · Control group 3: 23 enrolled, completions not described Age: 71 ± 6.7 years Sex: 88 female, 0 male Other medicines: · Vitamin D 1000IU daily · Women with low calcium intake (not defined) were administered calcium supplementation in a dose sufficient to achieve 1000mg/daily Inclusion criteria: · Continuous alendronate therapy for at least five years (10mg/day) · Post-menopausal women Exclusion criteria: · Not mentioned |
| Interventions | Medicine: Alendronate 10mg daily
Withdrawal Schedule: Not described Comparator: Group One: Alendronate deprescribed after five years continuous treatment Group Two: Alendronate deprescribed after at least one year and no more than four years continuous treatment Group Three: recently-diagnosed and untreated, post-menopausal osteoporotic patients |
| Outcomes | Bone Mass Density
Fractures Bone turnover markers Parathyroid and calcium and vitamin D levels |
| Dates | Dates: April 2006 to June 2007
Follow-up duration: 12 months |
| Funding sources | Not stated |
| Notes | Vitamin D levels assessed but no mention of what time of year, or if a consistent time of year for each participant (says baseline and 12 months – but Not described if all discontinued simultaneously or as a rolling process).
Authors stated: “Acknowledgments and conflict of interest disclosure: Marise Lazaretti-Castro is consultant from Sanofi-Aventis and Novartis, and participates as principal investigator in clinical research trials supported by Merck, Sharp & Dohme, Eli Lilly and Pfizer. C-terminal telopeptide (CTX) and P1NP kits were kindly provided by Roche Diagnósticos. There are no more potential conflicts of interest.” Note: for the purposes of analysis in the systematic review, the data from groups 1 and 2 have been used. |
Risk of bias table
| Bias | Authors’ judgment | Support for judgment |
| Random sequence generation (selection bias) | High risk | Not a randomized study – cohort study |
| Allocation concealment (selection bias) | High risk | Allocation was not concealed. Allocation was based on consecutive patients. |
| Blinding of participants and personnel (performance bias) | 5 out of 5 | The study was un-blinded. |
| Blinding of outcome assessment (detection bias) | 5 out of 5 | As above |
| Incomplete outcome data (attrition bias) | 5 out of 5 | Reasons for attrition were not given. Number of drop-outs from group 3 was not available. Missing outcome data for groups 2 and 3 for some outcomes, and no numbers provided for other outcomes (p-values given but no numbers to state effect size).
Results indicate a sizable proportion of drop-outs, but this was not addressed in the article. |
| Selective reporting (reporting bias) | 5 out of 5 | The study was not detailed in its explanation of pre-specified outcomes, and it only reported the outcomes for some groups. Values were often not given and only reported in relative terms or in p-values, so the clinical relevance is unclear.
CTX and Procollagen Type 1 N-Terminal Propeptide (PINP) were used to follow bone turnover – reported in graph |
| Confounding (non-randomized) | 3 out of 5 | No list of confounders but I could not find any serious deficiencies. The groups were actually paired for age and taken from a naturalistic setting |
| Other bias | High risk | Vitamin D levels measured at baseline and 12 months – unclear if rolling or simultaneous enrollment as the seasonal variations potentially mean that this data is non-comparable between participants and groups.
Industry funded – role of funder not described. Marise Lazaretti-Castro is consultant from Sanofi-Aventis and Novartis, and participates as principal investigator in clinical research trials CTX and P1NP kits were kindly provided by Roche Diagnósticos |
| Selection bias (Newcastle-Ottawa scale) | Unclear risk | The intervention group was a representative sample of post-menopausal women (age 71.0 ± 6.7 years old) on continuous alendronate therapy for at least five years.
The concurrent control cohort appears to have been drawn from the same population as the intervention cohort. Yes, they demonstrated that the outcome of interest was not present at the start of the study. Ascertainment of exposure was by secure record. |
| Comparability bias (Newcastle-Ottawa scale) | High risk | The study controlled for diagnosis of osteoporosis, rather than use of alendronate withdrawal. |
| Outcome bias (Newcastle-Ottawa scale) | Unclear risk | Outcome assessment by record linkage.
Follow-upduration was long enough for outcomes to occur. One or more years – time for clinically-relevant outcomes (e.g. fractures) to occur. Longer follow-upwould have been preferable, however, due to the long half-life of alendronate. Adequacy of follow-upuncertain. 12.5% may or may not be adequate, but no description is given to determine if this is the case |
Leave a Reply