Systematic Review: Comparative Effectiveness and Harms of Combinations of Lipid-Modifying Agents and High-Dose Statin Monotherapy

  1. Mukul Sharma, MD, MSc;
  2. Mohammed T. Ansari, MBBS, MMedSc, MPhil;
  3. Ahmed M. Abou-Setta, MD, PhD;
  4. Karla Soares-Weiser, MD, PhD;
  5. Teik Chye Ooi, MBBS;
  6. Margaret Sears, PhD;
  7. Fatemeh Yazdi, MSc;
  8. Alexander Tsertsvadze, MD, MSc; and
  9. David Moher, PhD
  1. From the University of Ottawa, Clinical Epidemiology Program, Ottawa Hospital Research Institute, and Children's Hospital of Eastern Ontario Research Institute, Ottawa, Ontario, Canada; Alberta Research Centre for Child Health Evidence, University of Alberta, Edmonton, Alberta, Canada; and Enhance Reviews, Kfar-Saba, Israel.

    Abstract

    Background: Statin therapy effectively prevents vascular disease, but treatment targets are often not achieved.

    Purpose: To compare the benefits and harms of high-dose statin monotherapy with those of combination therapy in adults at high risk for coronary disease.

    Data Sources: English-language records from MEDLINE (1966–2009), EMBASE (1980–2009), and the Cochrane Library (third quarter of 2008).

    Study Selection: A reviewer screened records, and a second reviewer verified selection of randomized, controlled trials in adult patients that compared combinations of statins and bile-acid sequestrants, fibrates, ezetimibe, niacin, or ω-3 fatty acids with statin monotherapy, as well as nonrandomized comparative studies that were longer than 24 weeks and reported clinical and harms outcomes.

    Data Extraction: Data were abstracted for studies by using standardized forms, and study quality was rated with a standardized scale and strength of evidence by using the Grading of Recommendations Assessment, Development, and Evaluation approach.

    Data Synthesis: 102 studies met eligibility criteria. The main analysis compared combination therapy with high-dose statin monotherapy in high-risk patients. Very-low-strength evidence showed that statin–ezetimibe (2 trials; n = 439) and statin–fibrate (1 trial; n = 166) combinations did not reduce mortality more than high-dose statin monotherapy. No trials compared the effect of combination therapy versus high-dose statin monotherapy on the incidence of myocardial infarction, stroke, or revascularization procedures. Two statin–ezetimibe trials (n = 295) demonstrated higher low-density lipoprotein cholesterol goal attainment with combination therapy (odds ratio, 7.21 [95% CI, 4.30 to 12.08]). Trials in lower-risk patients did not show a difference in mortality.

    Limitations: Studies were generally short, focused on surrogate outcomes, and were heterogeneous in study sample risk for coronary disease. Few studies examined treatment combinations other than statin–ezetimibe.

    Conclusion: Limited evidence suggests that combinations of lipid-lowering agents do not improve clinical outcomes more than high-dose statin monotherapy. Very-low-quality evidence favors statin–ezetimibe treatment for attainment of low-density lipoprotein cholesterol goals.

    Primary Funding Source: Agency for Healthcare Research and Quality.

    Editors' Notes

    Context

    • Do combinations of lipid-lowering agents improve outcomes more than statin monotherapy in adults with dyslipidemia?

    Contribution

    • This review of 102 studies found 2 trials that suggested lower target lipid levels were more often achieved with a statin–ezetimibe combination therapy than with high-dose statin monotherapy. No firm trial evidence showed that combining a statin with another agent (bile-acid sequestrant, fibrate, ezetimibe, niacin, or ω-3 fatty acids) improved clinical outcomes (myocardial infarction, stroke, or mortality) more often than high-dose statin monotherapy.

    Caution

    • Most trials were short in duration, focused on surrogate outcomes, and used similar doses of statins in the combination and monotherapy groups.

    —The Editors

    More than 28 million Americans have some form of cardiovascular disease (1). Resulting medical expenditures and lost productivity cost an estimated $431.8 billion in the United States in 2007 (2). Lowering low-density lipoprotein (LDL) cholesterol levels reduces rates of coronary heart disease (CHD) and ischemic stroke (3, 4). Successive versions of guidelines have established treatment thresholds and recommended aggressive treatment for high-risk persons (5). However, only one third of all patients, and proportionally fewer patients with established CHD, achieve guideline targets (6). Effective strategies in persons requiring intensive lipid-lowering therapy are critically needed.

    After dietary and lifestyle recommendations, statins are first-line medications for lipid-lowering therapy. These agents are used alone or in combination with other lipid-lowering medications. Treatment options for patients requiring intensive lowering of cholesterol levels include an increased dose of a statin medication alone or the use of a statin in combination with a lipid-modifying agent of another class. Ezetimibe, niacin, bile-acid sequestrants, fibrates, and ω-3 fatty acids are available treatment options for combination with statins (7–13). The Agency for Healthcare Research and Quality (AHRQ) commissioned a comparative effectiveness review to evaluate which of these strategies is superior with respect to clinical outcomes. In our review, we compare the benefits and risks of high-dose statin monotherapy with those of combination therapy for clinical events, surrogate measures, tolerability, and adherence in persons requiring intensive lipid-lowering therapy.

    Methods

    We address the following 2 key questions in this review:

    1. For patients who require intensive lipid-modifying therapy, what are the comparative long-term benefits and rates of serious adverse events of coadministration of different lipid-modifying agents (that is, a statin plus another lipid-modifying agent) compared with those of higher-dose statin monotherapy?

    2. Do these regimens differ in achievement of LDL cholesterol targets (or other surrogate markers), short-term side effects, tolerability, or adherence?

    We followed a standard protocol for this review. The full technical report (14), contains a detailed description of the methods and results, including search strategies and additional evidence tables. We expected that few reports would directly address a high-risk patient sample and compare combination therapy with high-dose statin monotherapy. In discussion with the AHRQ, we expanded the scope of patient risk and comparator dose. Our review includes studies that enrolled mixed-risk patients and used various statin doses.

    Data Sources and Searches

    We searched MEDLINE from 1966 to May 2009, EMBASE from 1980 to May 2009, and the Cochrane Library to the third quarter of 2008. We searched Scopus for references that cited 8 expert-nominated articles, the U.S. Food and Drug Administration statistical and medical reviews of drug applications, and the Internet. Information on published and unpublished studies of drugs was requested from Abbott, AstraZeneca, and Merck/Schering-Plough Pharmaceuticals by the Oregon Evidence-based Practice Center's Scientific Resource Center. We contacted corresponding authors of included studies for additional unpublished data relevant to the key questions.

    Study Selection

    We included English-language studies in adult patients comparing combinations of statins plus bile-acid sequestrants, fibrates, ezetimibe, niacin, or ω-3 fatty acids with statin monotherapy. We included randomized, controlled trials that reported mortality rates, vascular events, and lipid levels. Because we expected few long-term randomized trials, we also included long-term (≥24 weeks), nonrandomized comparative studies that reported clinical outcomes, serious adverse events, and cancer incidence. We excluded abstracts and reports on investigational agents.

    Data Extraction and Quality Assessment

    One reviewer screened record titles or abstracts to include studies; exclusions were independently verified by a second reviewer. Two reviewers independently screened full-text reports, with conflicts resolved by consensus. Data for the longest study follow-up were extracted in standardized forms. We assessed study quality as good, fair, or poor by predefined criteria. We also evaluated the strength of the body of evidence for all-cause mortality, vascular death, Adult Treatment Panel III (ATP III) goal attainment, and serious adverse events by using the GRADE (Grading of Recommendations Assessment, Development, and Evaluation) approach (15).

    Data Synthesis and Analysis

    All-cause mortality and vascular death were the main outcomes of interest. Other clinical outcomes were myocardial infarction, the acute coronary syndrome, stroke, transient ischemic attack, and revascularization procedures. We considered the following surrogate outcomes: attainment of ATP III LDL cholesterol goals, LDL cholesterol and high-density lipoprotein (HDL) cholesterol levels, and measures of carotid or coronary atherosclerosis. Harms outcomes were serious adverse events, cancer, withdrawals because of adverse events, incidence of at least 1 adverse event, elevated serum aminotransferase levels, hepatitis, myalgia, creatine kinase levels greater than 10 times the upper limit of normal, and rhabdomyolysis. Treatment adherence was also investigated.

    The main analyses were conducted in patients who required intensive lipid-lowering therapy, in which the statin in combination therapy was compared with a high dose of the same statin as monotherapy (Appendix Table 1). Patients requiring intensive lipid-lowering therapy included those with a 10-year CHD risk greater than 20%, mean baseline LDL cholesterol levels of at least 5.0 mmol/L (≥190 mg/dL), or both. We expected few reports addressing this specific dose comparison, so we conducted additional analyses unrestricted by patient risk and statin type and dose.

    Appendix Table 1. Definitions of Low and High Doses of Statins

    We used the DerSimonian and Laird approach, or Peto odds ratio for rare events, for meta-analysis. We used the Comprehensive Meta-Analysis software, version 2.2046 (Biostat, Englewood, New Jersey). We performed a meta-analysis when the set of trials did not have substantial heterogeneity. We avoided double counting for trials with multiple unequal numbers of treatment groups (14). We examined all-cause mortality, vascular death, and surrogate efficacy outcomes for patients who needed intensive lipid-lowering therapy, as well as for all patients. Evidence syntheses of harms and clinical outcomes other than all-cause mortality and vascular death were done regardless of patient risk. We defined long-term trials as those lasting longer than 24 weeks.

    Role of Funding Source

    The review was funded by the AHRQ. Staff from the AHRQ participated in the formulation of the research questions and reviewed the methods and the draft report. The funding source was not involved in study selection, data abstraction, and synthesis of results or approval of the manuscript for publication.

    Results

    We screened 9735 records and reviewed 923 full-text articles. The initial search was performed in August 2008 and updated in May 2009. When updating the evidence, we restricted inclusion to clinical efficacy outcomes, serious adverse events, and cancer incidence from studies of at least 24 weeks' duration. We included 98 unique randomized, controlled trials (16–113) and 4 nonrandomized studies (114–117) (Figure 1 and Appendix Table 2).

    Figure 1.
    View larger version:
    Figure 1. Literature search and selection.

    FDA = U.S. Food and Drug Administration; NRS = nonrandomized study; RCT = randomized, controlled trial.

    * Total does not sum to 48 because 1 study was excluded in 2 categories.

    Appendix Table 2. Distribution of Included Studies

    Characteristics of Included Studies

    Few studies compared combination therapy with high-dose statin monotherapy particularly in patients requiring intensive lipid-lowering therapy. Most studies were of fair quality, used strict eligibility criteria, excluded very sick patients, and compared similar doses of statins in combination and monotherapy, focusing on surrogate outcomes over a short-term period. Few clinical events were reported in the nonrandomized studies. Results are presented by outcome, the 10-year CHD risk stratum of the trial sample, and the statin dose in the monotherapy group. The Table summarizes the overall strength of evidence for the important outcomes.

    Table. Strength of Evidence Comparing Statins in Combination Therapy with High-Dose Statin Monotherapy in Patients Requiring Intensive Treatment

    Mortality

    Three (fair- to poor-quality) randomized, controlled trials comparing combinations of statin–ezetimibe (19, 108) and statin–fibrate (24) with high-dose statin monotherapy reported all-cause mortality in high-risk patients, but none specified vascular death. Mortality was rare, and no difference between treatments was noted.

    Meta-analysis of 14 short-term fair-quality trials (n = 6275) that used various doses and types of statins in combination with ezetimibe compared with statin monotherapy in high-risk patients demonstrated no statistically significant difference (but a wide confidence bound) in mortality (odds ratio, 0.61 [95% CI, 0.22 to 1.71]) (19, 30–32, 34, 81, 83, 84, 86, 88, 95, 106, 108, 112). For an analysis unrestricted by patient sample risk and dose, we meta-analyzed 3 fair-quality trials with more than 18 000 participants for the statin–ω-3 fatty acids combination and found no statistically significant difference in mortality (odds ratio, 1.08 [CI, 0.91 to 1.28]) (63, 82, 90).

    Other Clinical Outcomes

    No trials comparing a combination with high-dose statin monotherapy reported the occurrence of myocardial infarction, stroke, transient ischemic attacks, or revascularization procedures in patients requiring intensive lipid-lowering therapy. Rare events in the very few trials of statin–ezetimibe, statin–fibrate, statin–niacin, and statin–bile-acid sequestrant combinations lasting 12 to 52 weeks precluded meaningful conclusions for other clinical outcomes. One fair-quality, large statin–ω-3 fatty acids trial that used various statins and doses in mixed-risk patients reported no statistically significant difference between treatments for the outcomes of nonfatal myocardial infarction and stroke over 5 years (82). Our meta-analysis of 2 statin–ω-3 fatty acids trials demonstrated no statistically significant difference (but a wide confidence bound) between treatments for the outcome of fatal myocardial infarction (odds ratio, 0.73 [CI, 0.34 to 1.58]) (59, 82).

    Serious Adverse Events

    In 3 short-term fair-quality trials (n = 927) of the statin–ezetimibe combination compared with high-dose statin monotherapy in diverse patient samples, few (5%) participants had a serious adverse event (odds ratio for difference between treatments, 1.64 [CI, 0.85 to 3.19]) (16, 19, 21). Meta-analysis of 27 fair-quality trials, unrestricted by statin dose, showed no statistically significant difference in the rate of adverse events (odds ratio, 1.08 [CI, 0.88 to 1.33]) (16–22, 26–28, 30–32, 34, 36, 41, 48, 81, 83, 84, 86, 88, 93, 106, 108, 109, 112). For combinations of statin–fibrate (37, 39), statin–niacin (100, 103, 104, 107, 111), statin–bile-acid sequestrant (60, 61), and statin–ω-3 fatty acids (90), the evidence was restricted to few trials in mixed-risk patient samples that used similar statin doses in both groups, with absolute rates of serious adverse events between 2% and 3%.

    Cancer Incidence

    Evidence was limited to trials in mixed-risk patient populations with various statins and doses. A single fair-quality, 5-year, statin–ω-3 fatty acids trial, with 3% incidence of cancer, reported no statistically significant difference between treatments (82). Two 24- to 48-week statin–ezetimibe trials (n = 971), with 1% incidence of cancer, demonstrated no statistically significant difference between treatments (18, 28). Events were rare in a single small, statin–niacin trial (104).

    Attainment of ATP III LDL Cholesterol Goals

    Meta-analysis of 2 fair-quality trials that compared statin–ezetimibe trials with high-dose statin monotherapy in patients requiring intensive lipid-lowering therapy demonstrated a greater probability of goal attainment with combination therapy (odds ratio, 7.21 [CI, 4.30 to 12.08]) (19, 108). Twenty-three fair-quality trials of statin–ezetimibe combinations, unrestricted by statin dose, in diverse patient samples reported attainment of guideline LDL cholesterol goals (16, 17, 19, 27, 30–34, 36, 41, 48, 49, 51, 81, 83, 86, 88, 93, 95, 101, 106, 108). All but 1 of these trials reported a higher proportion of participants attaining goals with statin–ezetimibe combination treatment.

    LDL Cholesterol Levels

    Two fair-quality statin–ezetimibe trials assessed reductions in LDL cholesterol levels in patients requiring intensive lipid-lowering therapy. Both showed additional reductions of 10% to 20% with combination therapy (16, 19). Eighteen fair-quality statin–ezetimibe trials (16, 19, 20, 27, 30–32, 34, 36, 55, 81, 83, 84, 86, 88, 95, 99, 112) and 4 poor-quality statin–bile-acid sequestrant trials (38, 61, 76, 97) in high-risk patients used various statin doses in the monotherapy groups. All statin–ezetimibe trials favored combination treatment with mean additional LDL cholesterol reductions of 4% to 27%. Inconsistent results were found in the statin–bile-acid sequestrant trials.

    When lower doses of statins in combination therapies were compared with high doses of the same statin monotherapies across a wide spectrum of patient populations, additional LDL cholesterol reductions of 3% to 20% were demonstrated in 6 fair-quality statin–ezetimibe trials (Figure 2) (16–21). Most of the 35 fair-quality statin–ezetimibe trials unrestricted by patient sample risk and statin dose showed 4% to 27% additional reductions in LDL cholesterol levels with combination therapy (16–22, 26–28, 30–34, 36, 41, 42, 45, 48, 49, 51, 55, 80, 81, 83, 84, 86–88, 93, 95, 99, 109, 112). Eleven, mostly poor-quality, statin–bile-acid sequestrant trials could not be pooled because of heterogeneity. These trials were unrestricted by patient sample risk or statin dose and favored combination therapy (23, 38, 57, 60, 61, 71, 72, 74, 76, 97, 102). In 2 fair-quality statin-ω-3 fatty acids trials, unrestricted by dose and patient sample risk, statin monotherapy was superior (mean difference in LDL cholesterol reduction, 5.26% [CI, 1.79% to 8.74%]) (54, 90). The few statin–fibrate (24, 37, 89) and statin–niacin (25, 46, 68, 80, 103, 104, 110) trials showed inconsistent results.

    Figure 2.
    View larger version:
    Figure 2. Percentage of change in low-density lipoprotein cholesterol levels from baseline after low-dose statin combination therapy versus high-dose statin monotherapy in diverse populations.

    BAS = bile-acid sequestrant.

    HDL Cholesterol Levels

    Meta-analysis of 5 fair-quality randomized trials in diverse patient samples showed no statistically significant difference in percentage change in HDL cholesterol levels for the statin–ezetimibe combination versus high doses of statin alone (mean difference, 0.31 [CI, −0.89 to 1.52]) (17–21). Evidence for statin–fibrate and statin–niacin combinations for this comparison was insufficient.

    Measures of Atherosclerosis

    One trial of 642 evaluable participants compared simvastatin plus ezetimibe with identical-dose simvastatin monotherapy in participants requiring intensive lipid-lowering therapy. The mean between-group difference of change in mean carotid intima–media thickness was 0.01 mm [CI, −0.01 to 0.02 mm]) (99). Another trial in 149 evaluable participants requiring intensive lipid-lowering therapy and using various statins in combination with niacin reported no statistically significant difference in intima–media thickness (mean difference, −0.03 mm [CI, −0.06 to 0.003 mm]) (40).

    Treatment Adherence and Harm

    Evidence comparing a combination therapy with high-dose statin monotherapy was available only for the statin–ezetimibe combination in 5 short-term trials (16–19, 21), which showed no statistically significant differences between treatments in withdrawal because of adverse events and aminotransferase levels greater than 3 times the upper limit of normal.

    Meta-analysis of 10 trials evaluating various statins and doses in diverse patient samples showed that rates of early withdrawal because of adverse events were higher for the combination of statin plus niacin than for statin monotherapy (odds ratio, 2.38 [CI, 1.63 to 3.47]) (40, 46, 47, 50, 67, 100, 103, 104, 107, 111). More participants developed at least 1 adverse event with the statin–bile-acid sequestrant combination (odds ratio, 2.19 [CI, 1.28 to 3.75]) (38, 60, 71, 79). Rates of participants developing aminotransferase level elevations, hepatitis, increased creatine kinase levels, or myalgia did not differ between any combination therapy and statin monotherapy (14). In addition, no participant developed rhabdomyolysis across 27 trials, 85% of which were short-term and investigated all 5 combination therapies.

    No statistical difference in treatment adherence was observed for statin–ezetimibe (17, 18, 28, 31, 41, 45, 49, 80, 83, 84, 99, 106) and statin–niacin (25, 40, 46, 47, 80) combination therapies, whereas 5 trials of statin–bile-acid sequestrant combination therapy demonstrated inconsistent effects precluding meta-analysis (23, 38, 62, 79, 102).

    Discussion

    We found no benefit of combination therapy for mortality, myocardial infarction, stroke, and revascularization procedures over high-dose statin monotherapy in patients requiring intensive lipid-lowering therapy. Most studies were short and focused on surrogate markers, limiting their ability to detect any potential differences in important clinical outcomes. Assessment of these outcomes requires large long-term trials in high-risk patients. Data are insufficient to comment on any effects in racial or ethnic subgroups, women, and frail elderly persons. Future trials should include people from those and other specific populations of interest.

    The statin dose in most trials was similar in the combination and monotherapy groups. More intensive statin therapy is associated with a greater reduction in LDL cholesterol levels and vascular events without increasing discontinuation because of drug-related harms (118). Therefore, the appropriate comparator for combination therapies may be high-dose statin therapy.

    Very few studies examined combinations other than with statin–ezetimibe. Overall, limited evidence demonstrated a greater reduction in LDL cholesterol levels with any of the 5 combinations than with high–dose statin monotherapy in participants requiring intensive lipid-lowering therapy. In 2 trials, 10% to 20% additional reductions in LDL cholesterol were demonstrated in high-risk participants using low-dose simvastatin–ezetimibe combination therapy compared with high-dose statin monotherapy. With statin–ω-3 fatty acids combination therapy, participants in the monotherapy groups achieved greater reductions in LDL cholesterol levels. ω-3 fatty acids are primarily used to improve the lipid profile by decreasing triglyceride levels; elevations in LDL cholesterol levels with ω-3 fatty acids therapy are consistent with our results (13).

    The absence of an effect on carotid intima–media thickness despite reductions in LDL cholesterol levels was the unexpected outcome of 2 trials. Many explanations have been suggested, but this finding may be the result of the relatively short study period, because no intervention studies with a 2-year duration or less have demonstrated an effect on carotid intima–media thickness (119).

    Adverse events were not reported systematically, and many trial protocols did not specify standard methods for ascertainment of harms. The follow-up periods for most studies were too short to discern events with a long latency, such as cancer.

    Medication adherence is an important issue in determining population benefit. Less complex regimens using fewer separate agents may be preferable (120). Adherence was rarely reported, and most reports mentioned only data regarding the proportion of participants who discontinued treatment. Absolute rates of withdrawal in statin–niacin combination therapy in 4 trials were less than 10%, favoring statin monotherapy. In single trials, significantly fewer participants adhered to statin–bile-acid sequestrant and statin–niacin combination treatments. Both treatments have well-known short-term side effects, which are the probable cause of these findings (10).

    Our review has several limitations. We did not examine all possible combination medications, doses, or patient populations. Combinations may prove to be useful for selected lipid profiles or in patients who do not reach targets despite maximal therapy. The assessments of clinical outcomes, harms, and treatment adherence were limited by the shortage of long-term studies, with relatively few studies for combination therapies other than statin–ezetimibe and, for surrogate outcomes, by the statistical limitations of the outcome measures. The extent to which short-term surrogate measures correlate with outcome has been well demonstrated for statin monotherapy but remains unclear for combination therapies. We did not attempt imputation, because the possibility of double counting could not be avoided, limiting inclusion of composite outcomes. Our statistically conservative approach to meta-analyses precluded analyses in many instances. We rated the strength of the evidence as very low for the main outcomes (see Table) because of the lack of reports directly addressing the specified patient samples and comparators. The available evidence supporting the use of combination therapies over high-dose statin monotherapy, including long-term clinical benefits and reduced risks, is insufficient to guide many clinical decisions. The effectiveness of statins in reducing vascular event suggests that the benefits of additional therapies need to be clearly defined along with attendant risks and costs before advocating widespread use of combination treatment.

    Article and Author Information

    • Acknowledgment: The authors thank Margaret Sampson, Chantelle Garritty, Heather Clark, Robert Côté, Nick Barrowman, Raymond Daniel, and Sophia Tsouros for their collective efforts.

    • Grant Support: By the Agency for Healthcare Research and Quality, U.S. Department of Health and Human Services (contract 290-02-0021).

    • Potential Conflicts of Interest: Honoraria: M. Sharma (Merck & Co., Schering-Plough), T.C. Ooi (Oryx Pharmaceuticals, Fournier Pharma, AstraZeneca, Merck Frosst, Solvay Pharma, Schering-Plough).

    • Requests for Single Reprints: Mukul Sharma, MD, MSc, Canadian Stroke Network, Regional Stroke Program, The Ottawa Hospital, Civic Campus, C2, Room 2182, 1053 Carling Avenue, Ottawa, Ontario K1Y 4E9, Canada.

    • Current Author Addresses: Dr. Sharma: Canadian Stroke Network, Regional Stroke Program, The Ottawa Hospital, Civic Campus, C2, Room 2182, 1053 Carling Avenue, Ottawa, Ontario K1Y 4E9, Canada.

    • Drs. Ansari and Tsertsvadze and Ms. Yazdi: University of Ottawa Evidence-based Practice Center, CHEO-RI, 401 Smyth Road, Ottawa, Ontario K1H 8LI, Canada.

    • Dr. Abou-Setta: University of Alberta Evidence-based Practice Center, Alberta Research Centre for Health Evidence, Aberhart Centre One, Room 8412, 11402 University Avenue, Edmonton, Alberta T6G 2J3, Canada.

    • Dr. Soares-Weiser: Enhance Reviews, PO Box 137, Kfar-Saba, 44101, Israel.

    • Dr. Ooi: Division of Endocrinology and Metabolism, University of Ottawa, The Ottawa Hospital, Riverside Campus, 1967 Riverside Drive, Ottawa, Ontario K1H 7W9, Canada.

    • Dr. Sears: RR 1, Box 9012, Dunrobin, Ontario K0A 1T0, Canada.

    • Dr. Moher: University of Ottawa Evidence-based Practice Center, University of Ottawa, and Ottawa Methods Centre, Clinical Epidemiology Program, Ottawa Hospital Research Institute, 501 Smyth Road, 6th Floor, Critical Care Wing, Room W6112, Ottawa, Ontario K1H 8L6, Canada.

    • Author Contributions: Conception and design: M. Sharma, M.T. Ansari, A.M. Abou-Setta, T.C. Ooi, A. Tsertsvadze, D. Moher.

    • Analysis and interpretation of the data: M. Sharma, M.T. Ansari, A.M. Abou-Setta, K. Soares-Weiser, T.C. Ooi, M. Sears, A. Tsertsvadze, D. Moher.

    • Drafting of the article: M. Sharma, M.T. Ansari, A.M. Abou-Setta, M. Sears, A. Tsertsvadze, D. Moher.

    • Critical revision of the article for important intellectual content: M. Sharma, M.T. Ansari, A.M. Abou-Setta, T.C. Ooi, M. Sears, A. Tsertsvadze, D. Moher.

    • Final approval of the article: M. Sharma, M.T. Ansari, A.M. Abou-Setta, K. Soares-Weiser, T.C. Ooi, M. Sears, D. Moher.

    • Provision of study materials or patients: F. Yazdi.

    • Statistical expertise: M.T. Ansari, A.M. Abou-Setta, D. Moher.

    • Obtaining of funding: D. Moher.

    • Administrative, technical, or logistic support: M.T. Ansari, A.M. Abou-Setta, M. Sears, F. Yazdi, D. Moher.

    • Collection and assembly of data: M. Sharma, M.T. Ansari, A.M. Abou-Setta, K. Soares-Weiser, T.C. Ooi, M. Sears, F. Yazdi, A. Tsertsvadze.

    References

    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    11. 11.
    12. 12.
    13. 13.
    14. 14.
    15. 15.
    16. 16.
    17. 17.
    18. 18.
    19. 19.
    20. 20.
    21. 21.
    22. 22.
    23. 23.
    24. 24.
    25. 25.
    26. 26.
    27. 27.
    28. 28.
    29. 29.
    30. 30.
    31. 31.
    32. 32.
    33. 33.
    34. 34.
    35. 35.
    36. 36.
    37. 37.
    38. 38.
    39. 39.
    40. 40.
    41. 41.
    42. 42.
    43. 43.
    44. 44.
    45. 45.
    46. 46.
    47. 47.
    48. 48.
    49. 49.
    50. 50.
    51. 51.
    52. 52.
    53. 53.
    54. 54.
    55. 55.
    56. 56.
    57. 57.
    58. 58.
    59. 59.
    60. 60.
    61. 61.
    62. 62.
    63. 63.
    64. 64.
    65. 65.
    66. 66.
    67. 67.
    68. 68.
    69. 69.
    70. 70.
    71. 71.
    72. 72.
    73. 73.
    74. 74.
    75. 75.
    76. 76.
    77. 77.
    78. 78.
    79. 79.
    80. 80.
    81. 81.
    82. 82.
    83. 83.
    84. 84.
    85. 85.
    86. 86.
    87. 87.
    88. 88.
    89. 89.
    90. 90.
    91. 91.
    92. 92.
    93. 93.
    94. 94.
    95. 95.
    96. 96.
    97. 97.
    98. 98.
    99. 99.
    100. 100.
    101. 101.
    102. 102.
    103. 103.
    104. 104.
    105. 105.
    106. 106.
    107. 107.
    108. 108.
    109. 109.
    110. 110.
    111. 111.
    112. 112.
    113. 113.
    114. 114.
    115. 115.
    116. 116.
    117. 117.
    118. 118.
    119. 119.
    120. 120.

    Responses to this article