Background
-
to identify and collect existing literature on the risks and benefits of use of metformin in the treatment of DM2 in older adults,
-
to assess the quality of the evidence identified, and develop recommendations when to discontinue or to adjust the dose of metformin in the treatment of DM2 in older adults.
Methods
Study inclusion criteria
Types of studies
Types of participants
-
overall mean or median age ≥ 65 years; or
-
overall mean or median age < 65 with subgroup analysis reporting on participants ≥65 years; or
-
overall mean or median age not reported but 80% or more of the included studies reporting a mean or median age ≥ 65 years.
-
≥80% of participants ≥65 years; or
-
<80% of participants ≥65 years with subgroup analysis reporting on participants ≥65 years.
Types of interventions
Types of outcomes
-
Quality of life
-
Mortality
-
Life expectancy
-
Hospitalisations
-
Cognitive impairment or cognitive status
-
Functional impairment or status
-
Cardiovascular event including stroke
-
Renal failure
-
Composite end points including any of the above (extraction of individual outcomes was undertaken if reported by original studies)
-
Adverse drug event including hypoglycaemia
-
Any of the above evaluated as safety endpoints.
Setting
Language
Search method
-
Search 1: Systematic reviews and meta-analyses in the Cochrane database of Systematic Reviews (2005 to 2015) and the Database of Abstracts or Reviews of Effects (1991 to 2015).
-
Search 2: Systematic reviews and meta-analyses in MEDLINE and MEDLINE (R) In-Process & Other Non-Indexed Citations (1946 to 2015), EMBASE (1974 to 2015), Health Technology Assessment (HTA) (2001 to 2015) and International Pharmaceutical Abstracts (IPA) (1970 to 2015).
-
Search 3A: Interventional and observational studies meeting eligibility criteria included in systematic reviews which did not meet the inclusion criteria for searches 1 and 2
-
Search 3B: Additional controlled interventional and observational studies identified from MEDLINE, MEDLINE (R) In-Process & Other Non-Indexed Citations, EMBASE, HTA and IPA published since 2011.
Data management
Selection of studies
Data extraction
Quality appraisal
Dealing with duplicate and companion publications
Data synthesis
Identification of “references of interest” for the development of recommendations
Development of recommendations
Results
Results of the search
Included studies
Authors and publication year | Type of study | Aim | Sample size and information about the amount of older participantsa
| Follow-up | Outcomes and measurement tools if applicable |
---|---|---|---|---|---|
Cryer 2005 [31] | Randomised, open label, parallel-group, multicentre, clinical trial | To evaluate the risk of lactic acidosis or other SAEs with metformin, under usual care conditions. | P: 8732
P ≥ 65 years: 3084 P using metformin (metformin only or in combination with SU or non-SU oral agent): 2515 P under usual care (SU, TZD, insulin, or any other non-metformin monotherapy or combination therapy): 569 | 12 months | Incidence of SAEsc, hospitalization, and death. |
Evans 2010 [32] | Population-based prospective cohort study | To examine the efficacy of metformin and SU in patients with DM2 and CHF. | P: 422
P ≥ 65 years: 365 P using SU: 191 P using any metformin (alone/combination): 174 | 1 year and at end of follow-up (death, loss to follow-up, or end of 10 year study window) | All-cause mortality |
Hung 2013 [33] | Population-based retrospective cohort study | To assess the risk of non-fatal cardiovascular events among patients with DM2 who are taking metformin monotherapy, glimepiride or glyburide. | P: 1159
P ≥ 71 years: 231 P using glyburide: 72 P using glimepiride: 50 P using metformin: 109 | 3 months | Incidence of non-fatal cardiovascular events including coronary artery disease, peripheral artery disease, stroke and heart failure. |
Inzucchi 2005 [34] | Retrospective cohort study | To determine the impact of insulin sensitizers on outcomes in diabetic patients after hospitalization with AMI. | P: 8872
P ≥ 65 years: 8872 P using no insulin sensitizer: 6641 P using metformin: 1273 P using TZD: 819 P using metformin + TZD: 139 | 1 year | Time from hospital discharge to death from any cause censored at 1 year of discharge. Time to first readmission for MI, first readmission for heart failure, and first readmission for any cause up to 1 year after discharge. |
Janka 2007 [35] | Parallel-group, open-label, randomized, multinational clinical trial | To investigate whether the safety and efficacy benefits of initiating insulin therapy with glargine and continued OADs, versus switching to premixed insulin, as previously reported, were also observed in the subset of patients aged 65 and older. | P: 130
P ≥ 65 years: 130 P using insulin glargine + glimepiride and metformin: 67 P using premixed insulin: 63 | 24 weeks | Hypoglycaemic events and their frequency. |
Josephkutty et al. 1990 [36] | Randomised double-blind cross-over study | Efficacy, metabolic effects, and acceptability of metformin were compared with tolbutamide in 20 diabetic patients aged between 65 and 95 years. | P: 20
P ≥ 65 years: 20 P using metformin/tolbutamide: 10 P using tolbutamide/metformin: 10 | 3 months with each treatment | Side-effects of drug treatment. |
Lapane et al. 2015 [37] | Retrospective cohort study | To evaluate the extent to which SU use was associated with fractures and falls among nursing home residents with DM2. | P: 11,958
P ≥ 65 years: 10,916 P using SU: 5128 P using biguanide: 5788 | Median 683 days (range: 1–1002 days) | Severe hypoglycaemia, falls, and fractures occurring in parts of the body typically associated with falling. |
MacDonald 2010 [38] | Case-control study | To examine outcomes in patients with DM2 and heart failure and to determine whether outcomes were associated with antidiabetic drug therapy. | P: 3266
P ≥ 65 years: 3102 Cases (P with DM2 and heart failure who died): 1633 Controls (P with DM2 and CHF alive): 1633 | Average 11 years | All-cause mortality. |
Masoudi 2005 [39] | Retrospective cohort study | To assess the relationship between the prescription of insulin-sensitizing agents (metformin and/or TZD) and death or readmission ofelderly diabetic patients initially admitted with heart failure in a cohort derived from the National Heart Care Project (NHC). | P: 5296
P ≥ 65 years: 5296 P using metformin: 1861 P using TZD: 2226 P using no insulin sensitizer: 12,069 | 1 year | Time from hospital discharge to death due to any cause, time to first readmission for any cause or for heart failure, proportions of patients who died or were readmitted at least once in the year after discharge, rates of readmission for the primary diagnosis of metabolic acidosis. |
Moore et al. 2013 [40] | Cross sectional study | To investigate the associations of metformin, serum vitamin B12, calcium supplements, and cognitive impairment in patients with DM2. | P: 1354
P ≥ 65 years: 1164 P with DM2/impaired glucose tolerance: 126 P using metformin: 35 P no using metformin: 91 | Not applicable | Cognitive performance measured with the Mini-Mental State Examination (MMSE). |
Roumie 2012 [41] | Retrospective cohort study | To compare the effects of SU and metformin monotherapy on CVD outcomes (AMI and stroke) or death. | P: 253,640
P ≥ 65 years: 118,014 P using metformin: 64,009 P using SU: 54,005 | 12 months | Hospitalization for AMI or stroke, or death. Composite of AMI and stroke events only. |
Roussel et al. 2010 [42] | Prospective, observational registry | To assess whether metformin use was associated with difference in mortality after adjustment for baseline differences and for the propensity to receive metformin among diabetics with established coronary artery disease, cerebrovascular disease, or peripheral arterial disease. | P: 19,553
P ≥ 65 years: 12,649 P using metformin: 4389 P without metformin: 8260 | 2 years | 2-year all-cause mortality. Cardiovascular death and first-occurring event among death, MI, or stroke. |
Schweizer 2009 et al. [43] | Randomized, active-controlled, parallel-group study | To compare the efficacy and tolerability of vildagliptin with metformin in elderly patients with DM2. | P: 335
P ≥ 65 years: 335 P using metformin: 166 P using vildagliptin: 169 | 24 weeks | AEs. |
Solomon 2009 [44] | Retrospective cohort study | To determine the association between TZD use and fracture risk among older adults with DM2. | P: 20,291
P ≥ 65 years: 20,291 P using TZD: 2347 P using SU: 13,709 P using metformin: 4235 | Follow-up ended at the first of any of the following events: death, loss of eligibility for Medicare or the drug benefit program, 180 days after the last dosage of oral hypoglycaemic agent, or end of follow-up. | Incidence of fracture within the cohort. |
Tzoulaki 2009 [45] | Retrospective cohort study | To investigate the risk of incident MI, congestive heart failure, and all-cause mortality associated with prescription of OADs monotherapies and combinations. | P: 91,521
P ≥ 65 years: 45,760 P using FGS: 4764b
P using SGS: 40,266 b
P using rosiglitazone: 4437b
P using rosiglitazone combination: 4642b
P using pioglitazone (alone/combination): 1879b
P using other drugs or combinations: 21,994b
P using metformin: 37,357b
| Mean follow-up 7.1 years | First occurrence of incident MI, CHF, and all-cause mortality. |
Study designs
Participants
Covariates in models
Interventions and outcomes
Main findings
Authors and publication year | Outcomes | Metformin cases/na (%) | Comparator cases/na (%) | Unadjusted Risk ratiob
(95% CI) | Reported Statistical comparisonc
| Result favours |
---|---|---|---|---|---|---|
Tier 1 outcomes (hypoglycaemia and adverse events): comparisons against other non-specific treatments | ||||||
Cryer 2005 [31] | Metformin | Usual care | ||||
Subgroup ≥ 65 years | Any SAE | 371/2515 (14.8) | 93/569 (16.3) | 0.90 (0.73, 1.11) | NR | M |
Tier 1 outcomes (hypoglycaemia and adverse events): comparisons against other specific treatments | ||||||
Janka 2007 [35] | Insulin glargine + OAD (glimepiride and metformin) | Premixed insulin | ||||
(n = 63 patients) | ||||||
Confirmed + unconfirmed hypoglycaemia | (n = 67 patients) | |||||
5.6 events per p-yr. | 11.4 events per p-yr | 0.49 (0.41, 0.59) |
p = 0.01 | M | ||
Confirmed hypoglycaemia | 3.7 events per p-yr | 9.1 events per p-yr |
p = 0.008 | M | ||
Confirmed symptomatic hypoglycaemia | 2.2 events per p-yr | 5.0 events per p-yr | 0.40 (0.33, 0.50) |
p = 0.06 | M | |
Confirmed nocturnal hypoglycaemia | 0.4 events per p-yr | 0.7 events per p-yr | 0.44 (0.33, 0.59) |
p = 0.26 | M | |
Severe hypoglycaemia | 0.0 events per p-yr | 0.1 events per p-yr |
p = 0.21 | M | ||
One or more treatment-emergent AEs | 32/67 (47.8) | 27/63 (42.9) | 0.55 (0.27, 1.12) | |||
0.14 (0.01, 2.61) | NR | C | ||||
1.11 (0.76, 1.63) | ||||||
Josephkutty et al. 1990 [36] | Metformin | Tolbutamide | NR | C | ||
Side effects | 32 side effects reported by 21 patients | 15 side effects reported by 20 patients | ||||
Schweizer et al. 2009 [43] | Metformin | Vildagliptin | ||||
AEs | 83/165 (50.3) | 74/167 (44.3) | 1.14 (0.90, 1.43) | NR | C | |
SAEs | 6/165 (3.6) | 5/167 (3.0) | 1.21 (0.38, 3.90) | NR | C | |
Gastrointestinal AEs | 41/165 (24.8) | 25/167 (15.0) | 1.66 (1.06, 2.60) | NR | C | |
Hypoglycaemia | 2/165 (1.2) | 0/167 (0.0) | 5.06 (0.24, 104.61) | NR | C | |
Tier 2 outcomes: comparisons against other non-specific treatments | ||||||
Cryer 2005 [31] | Metformin | Usual care | ||||
All-cause mortality | 60/2515 (2.4) | 12/569 (2.1) | 1.13 (0.61, 2.09) |
p = 0.878 | C | |
Subgroup ≥ 65 years | All-cause hospitalisations | 334/2515 (13.3) | 88/569 (15.5) | 0.86 (0.69, 1.07) |
p = 0.178 | M |
Inzucchi 2005 [34] | Metformin | No insulin sensitizer | ||||
1-year mortality | 246/1273 (19.3) | 2014/6641 (30.3) | 0.64 (0.57, 0.72) | HR = 0.92 (0.81, 1.06) | M | |
1-year MI readmission | 210/1273 (16.5) | 1247/6641 (18.8) | 0.88 (0.77, 1.00) | HR = 1.02 (0.86, 1.20) | C | |
1-year HF readmission | 435/1273 (34.2) | 2859/6641 (43.1) | 0.79 (0.73, 0.86) | HR = 1.06 (0.95, 1.18) | C | |
1-year all-cause readmission | 759/1273 (59.6) | 4268/6641 (64.3) | 0.93 (0.88, 0.97) | HR = 1.04 (0.96, 1.13) | C | |
Masoudi 2005 [39] | Metformin | No insulin sensitizer | ||||
Mortality | 460/1861 (24.7) | 4345/12069 (36.0) | 0.69 (0.63, 0.75) | HR = 0.87 (0.78, 0.97) | M | |
All-cause readmission | 1265/1861 (68.0) | 8702/12069 (72.1) | 0.94 (0.91, 0.97) | HR = 0.94 (0.89, 1.01) | M | |
HF readmission | 1091/1861 (58.6) | 7821/12069 (64.8) | 0.90 (0.87, 0.94) | HR = 0.92 (0.86, 0.99) | M | |
Readmission for metabolic acidosis | 2.3% | 2.6% |
P = 0.40 | |||
MacDonald 2010 [38] | Metformin | No antidiabetic drugs | ||||
Mortality | 155/376 (41) | 733/1306 (56) | 0.73 (0.65, 0.84) | OR = 0.65 (0.48, 0.87) | M | |
Moore et al. 2013 [40] | Metformin | Not on metformin | ||||
(n = 35 patients) | (n = 91 patients) | |||||
Cognitive performance | NR | NR | OR = 1.75 (0.81, 3.78) | C | ||
p = 0.158 | ||||||
Roussel et al. 2010 [42] | Mortality: | Metformin | No metformin | |||
Patients 65–80 years | 191/3791 (5.0) | 532/6768 (7.9) | 0.64 (0.55, 0.75) | HR = 0.77 (0.62, 0.95), p = 0.02 | M | |
Patients >80 years | 71/598 (11.9) | 220/1492 (14.7) | 0.81 (0.63, 1.03) | HR = 0.92 (0.66, 1.28), p = 0.61 | M | |
Tier 2 outcomes: comparisons against other specific treatments | ||||||
Evans 2010 [32] | Metformin monotherapy + combination (n = 205) | SU monotherapy | ||||
(n = 217) | ||||||
1-year mortality | NR | NR | OR = 0.60 (0.37, 0.97) | M | ||
Long-term mortality | NR | NR | OR = 0.67 (0.51, 0.88) | M | ||
Roumie 2012 [41] | Metformin | SU | ||||
(n = 64,009 patients) | (n = 54,005 patients) | |||||
Subgroup ≥ 65 years | Hospitalization for acute MI, stroke or death | 15.9 per 1,000p–yrs | 24.6 per 1,000p–yrs | HR = 0.85 (0.78, 0.92) | M | |
Hospitalization for acute MI or stroke | 12.9 per 1,000p–yrs | 18.5 per 1,000p–yrs | HR = 0.88 (0.81, 0.97) | M | ||
Lapane 2015 [37] | Hospitalisation for hypoglycaemia | Metformin monotherapy | SU monotherapy | |||
(n = 6151) | (n = 5807) | |||||
All ages | 132 in 6518 p-yrs | 289 in 6307 p-yrs | 0.43 (0.35, 0.53) | HR = 0.42 (0.33, 0.53) | M | |
Age 75–84 | 55 in 2524 p-yrs | 104 in 2455 p-yrs | 0.51 (0.37, 0.71) | HR = 0.50 (0.34, 0.73) | M | |
Age 85+ | 39 in 2248 p-yrs | 100 in 2167 p-yrs | 0.38 (0.26, 0.54) | HR = 0.38 (0.25, 0.58) | M | |
Hospitalisation for fractures related to falls | ||||||
All ages | 180 in 6305 p-yrs | 194 in 6174 p-yrs | 0.94 (0.76, 1.15) | HR = 0.88 (0.69, 1.12) | M | |
Age 75–84 | 70 in 2478 p-yrs | 86 in 2375 p-yrs | 0.78 (0.57, 1.07) | HR = 0.73 (0.50, 1.05) | M | |
Age 85+ | 74 in 2142 p-yrs | 65 in 2114 p-yrs | 1.12 (0.81, 1.57) | HR = 1.05 (0.68, 1.59) | C | |
Falls | ||||||
All ages | 1844 in 4546 p-yrs | 1864 in 4560 p-yrs | 0.99 (0.93, 1.06) | HR = 1.02 (0.94, 1.11) | C | |
Age 75–84 | 703 in 1785 p-yrs | 756 in 1693 p-yrs | 0.88 (0.80, 0.98) | HR = 0.90 (0.79, 1.02) | M | |
Age 85+ | 697 in 1519 p-yrs | 691 in 1547 p-yrs | 1.03 (0.92, 1.14) | HR = 0.98 (0.86, 1.12) | M | |
Hung 2013 [33] | Metformin (n = 109) | Glyburide (n = 72) | ||||
47 in 181 p-yrs | ||||||
Subgroup ≥ 71 years | Non-fatal CVD | 30 in 414 p-yrs | Glimepiride (n = 50) | 0.28 (0.18,0.44) | HR = 0.30 (0.18, 0.48) | M |
Non-fatal CVD | 30 in 414 p-yrs | 18 in 167 p-yrs | 0.67 (0.38, 1.21) | NR | M | |
Inzucchi 2005 [34] | Metformin | Thiazolidinedione | ||||
Mortality | 246/1273 (19.3) | 237/819 (28.9) | 0.67 (0.57, 0.78) | NR | M | |
MI readmission | 210/1273 (16.5) | 154/819 (18.8) | 0.88 (0.73, 1.06) | NR | M | |
HF readmission | 435/1273 (34.2) | 402/819 (49.1) | 0.70 (0.63, 0.77) | NR | M | |
All-cause readmission | 759/1273 (59.6) | 555/819 (67.8) | 0.88 (0.82, 0.94) | NR | M | |
Solomon 2009 [44] | Metformin | Thiazolidinediones | ||||
Fractures | 110/4235 (2.6) | 74/2347 (3.2) SU | 0.82 (0.62,1.10) | RR = 0.76 (0.56, 1.02) | M | |
Fractures | 110/4235 (2.6) | 480/13709 (3.5) | 0.74 (0.60, 0.91) | NR | M | |
Tzoulaki 2009 [45] | Metformin monotherapy | 1st generation SU monotherapy | ||||
>1.6mil intervals d
| ||||||
Subgroup ≥ 65 years | MI | NR | NR | HR = 0.79 (0.65, 0.96) | M | |
CHF | NR | NR | HR = 0.76 (0.68, 0.85) | M | ||
All-cause mortality | NR | NR | HR = 0.72 (0.67, 0.79) | M | ||
2nd generation SU monotherapy | ||||||
MI | NR | NR | HR = 0.82 (0.74, 0.91) | M | ||
CHF | NR | NR | HR = 0.85 (0.79, 0.91) | M | ||
All-cause mortality | NR | NR | HR = 0.74 (0.70, 0.78) | M | ||
Rosiglitazone monotherapy | ||||||
MI | NR | NR | HR = 0.85 (0.54, 1.33) | M | ||
CHF | NR | NR | HR = 0.93 (0.65, 1.33) | M | ||
All-cause mortality | NR | NR | HR = 0.98 (0.76, 1.27) | M | ||
Rosiglitazone combination | ||||||
MI | NR | NR | HR = 0.81 (0.63, 1.06) | M | ||
CHF | NR | NR | HR = 0.76 (0.61, 0.93) | M | ||
All-cause mortality | NR | NR | HR = 1.10 (0.94, 1.28) | C | ||
Pioglitazone alone and combined | ||||||
MI | NR | NR | HR = 1.23 (0.74, 2.08) | C | ||
CHF | NR | NR | HR = 0.90 (0.64, 1.26) | M | ||
All-cause mortality | NR | NR | HR = 1.54 (1.15, 2.04) | C | ||
Other drugs and combinations | ||||||
MI | NR | NR | HR = 0.87 (0.77, 0.98) | M | ||
CHF | NR | NR | HR = 0.93 (0.85, 1.01) | M | ||
All-cause mortality | NR | NR | HR = 0.74 (0.70, 0.78) | M |
Tier 1 outcomes (hypoglycaemia and adverse events)
Tier 2 outcomes: Metformin compared to other non-specific active treatments
Tier 2 outcomes: Metformin compared to other specific active treatments
Excluded studies
Quality appraisal of included studies
Randomised trials
Source | Type of study | Selection bias | Performance bias | Detection bias | Attrition bias | Reporting bias | ||
---|---|---|---|---|---|---|---|---|
1. Random sequence generation | 2. Allocation concealment | 3. Blinding of participants and personnel | 4. Blinding of outcome assessment | 5. Incomplete outcome data | 6. Selective reporting | 7. Other bias | ||
Cryer 2005 [31] | Randomised, open label, parallel-group, multicentre, clinical trial | UR | HR | HR | UR | LR | UR | HR |
Janka 2007 [35] | Parallel-group, open-label, randomized, multinational clinical trial | LR | LR | HR | HR | LR | LR | LR |
Josephkutty 1990 [36] | Randomized double-blind cross-over study | UR | UR | UR | UR | UR | UR | UR |
Schweizer 2009 [43] | Randomized, active-controlled, parallel-group study | UR | UR | UR | UR | LR | UR | HR |
Observational studies
Source | Type of study | 1. Focused issue | 2. Appropriate method | 3. Recruitment | 4. Selection of controls | 5. Exposure measured | 6. Outcome measured | 7. Confounding factors identified | 8. Confounding design/analysis | 9. Follow up complete | 10. Follow up long | 11. Results of this study | 12. How precise results/risk estimate | 13. Believe results | 14. Results be applied | 15. Results fit evidence |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Evans 2010 [32] | Population-based cohort study | Y | Y | Y | NA | Y | Y | N | N | Y | U | Y | Y | U | ||
Hung 2013 [33] | population-based retrospective cohort study | Y | U | Y | NA | Y | N | N | Y | U | U | Y | Y | Y | ||
Inzucchi 2005 [34] | Retrospective cohort study | Y | Y | Y | NA | N | Y | Y | Y | Y | N | U | Y | U | ||
Lapane 2015 [37] | Retrospective cohort study | Y | Y | Y | NA | Y | N | Y | Y | U | N | Y | Y | Y | ||
MacDonald 2010 [38] | Case-control study | Y | Y | Y | Y | Y | Y | Y | Y | Y | Y | Y | Y | Y | ||
Masoudi 2005 [39] | Retrospective cohort | Y | Y | Y | NA | Y | N | Y | Y | U | U | U | Y | U | ||
Moore 2013 [40] | Cross sectional study | Y | Y | N | N | N | Y | U | Y | NA | NA | U | U | N | ||
Roumie 2012 [41] | Retrospective cohort study | Y | Y | U | NA | U | U | Y | Y | U | U | U | Y | U | ||
Roussel 2010 [42] | Prospective, observational registry | Y | Y | N | NA | U | Y | U | Y | Y | U | Y | Y | Y | ||
Solomon 2009 [44] | Retrospective cohort | Y | Y | Y | NA | Y | Y | Y | Y | Y | N | U | Y | Y | ||
Tzoulaki 2009 [45] | Retrospective cohort study | Y | Y | Y | NA | U | Y | Y | Y | Y | Y | U | Y | Y |
Additional references of interest for the development of recommendations
Recommendations
Recommendations | Strength of the recommendation | Quality of the evidence | Type of evidence |
---|---|---|---|
It is suggested to discontinue metformin for the management of type 2 diabetes mellitus in patients with 2 or more of the following risk factors: age > 80; gastrointestinal complaints during the last year; GFR ≤60 ml/min. The benefit of metformin in this patient is uncertain and it is possibly outweighed by the risk of adverse drug reactions, depending on their severity. | Weak | Low | |
Reason: uncertainty about the magnitude of the benefits and harms. | It was considered to downgrade the quality of the evidence to low quality because there were study limitations (1 observational study with limitations and 2 RCTs with unclear risk of bias), indirectness (observational study with subgroup analysis), inconsistency (different types of comparisons evaluated). | ||
It is suggested to discontinue metformin for the management of type 2 diabetes mellitus in patients 80 years and older taking the life expectancy, physical and functional status of the patient into account. Patients who are concerned about adverse events or appear to experience AE may reasonably choose not to take metformin. | Weak | Low | Observational study [42] |
Reason: uncertainty about the magnitude of the benefits and harms. | It was considered to keep the quality of the evidence as low quality because this observational study had limitations: data in older people was from subgroup analysis, lack of reporting on recruitment and confounding factors. | ||
It is suggested to discontinue metformin for the management of type 2 diabetes mellitus in patients with gastrointestinal complaints taking the possible benefit and the severity of the patient complaints as possible dverse drug reactions into account. | Weak | Low | |
Reason: small RCTs with low quality and no significant benefits with metformin; uncertainty about the magnitude of the benefits and harms. | It was considered to downgrade the quality of the evidence to low quality because there were study limitations (2 RCTs with unclear risk of bias) and inconsistency (different types of comparisons evaluated). | ||
It is suggested to discontinue metformin for the management of type 2 diabetes mellitus in patients with renal insufficiency because metformin may increase the risk of lactic acidosis. | Weak | Low | Clinical guideline [46] |
Reason: evidence from a clinical guideline; uncertainty about the magnitude of the benefits and harms. | It was considered to keep the quality of the evidence as low quality because it was from a clinical guideline. |