Online supplement Table S1: Study characteristics and quality scores of single TIH case reports.

Transcription

1 Online supplement Table S1: Study characteristics and quality scores of single TIH case reports. No. Author et al. Year No. of TIH Study design Country Setting Hyponatremia Quality score Reference subjects definition 1 Achinger Case series USA Primary 104 mm Adrogue Review with case report USA Primary <136 mm Al-Salman Case report USA Secondary 115 mm Ayus Case report USA Secondary 116 mm Benfield Case report UK Secondary 124 mm Berl Case report USA Primary 96 mm Cakir Case report Turkey Secondary 119 mm Coler Case Report USA Secondary 120 mm Collier Case report UK Secondary 115 mm Cundy Case report UK Primary 99 mm Eastell Case report UK Primary 107 mm Fadel Case report and rechallenge Belgium Secondary 113 mm 5 27 study 13 Fuisz Case report and rechallenge USA Secondary 121 mm 4 28 study 14 Gardner Case report USA Secondary 110 mm Ghose Case report UK Primary 127 mm Gossain Case series USA Secondary <132 mm Hamburger Case report USA Primary 107 mm Handler Case report USA Secondary <135 mm

2 19 Husby Case report Denmark Primary 104 mm Hussain Case report Ireland Secondary 99 mm Jen Case report China Secondary <137 mm Johnson Case report USA Primary 108 mm Karp Retrospective study USA Primary 104 mm Kennedy Case report USA Secondary 108 mm Kone Case report USA Secondary 118 mm Lin Case report China Secondary 94 mm Luft Case report Germany Primary 115 mm Lundbom Case report Finland Secondary <116 mm Mataverde Case report USA Secondary 109 mm Menashe Case report Israel Primary 106 mm Meuleman Case report USA Primary 116 mm Miyasaka Case report Japan Secondary (Not specified) Mok Case report China Secondary Mouallem Cohort study Israel Secondary (Not specified) Mount Case vignette USA Secondary 113 mm Moussa Case report United Arab Secondary 110 mm 3 51 Emirates 37 Mozes Case report Israel Secondary 104 mm Onozaki Case report Japan Primary 124 mm Orija Case report USA Secondary <136 mm Ponte Case report USA Secondary 115 mm

3 41 Ranta Case study USA Primary 42 Roberts Case report UK Primary 43 Shah Case report (letter) USA Secondary 44 Sterns Case Report USA Secondary 45 Strykers Case report USA (Not specified) 46 van Assen Case report and rechallenge The Secondary study Netherland s 47 Wierzbicki Case report UK Primary 48 Wijnen Case report The Netherland s Secondary 49 Yap Case report China Secondary 125 mm mm mm mm mm mm mm mm mm 3 64 Table S2: Characteristics and quality scores of studies included in the meta-analysis i.e. where more than one TIH patient was reported per study. No. Author et al. Year No. of TIH subjects Study design Country Setting Hyponatremia definition Quality score Reference 1 Aaseth Case series Norway Primary <125 mm Adams Clinical survey UK Secondary <130 mm Al Qahtani Case series Saudi Arabia Secondary 135 mm Ambrosi Case reports France Secondary 116 mm Ashraf Case reports USA Primary 116 mm Bain Case report UK Secondary 112 mm Bayer AJ Case series UK Secondary <130 mm

4 8 Bissram Retrospective cohort USA Secondary <134 mm 7 72 study 9 Booker Case reports Australia Secondary <121 mm Canning G Case reports UK (Not specified) 125 mm Chapman MD Descriptive analysis Australia (Not specified) (Not specified) 3 75 using case reports 12 Chow Case series China Secondary <130 mm Clayton Cross sectional UK Primary <135 mm 4 77 observational study 14 Coenraad Case series The Netherlands Secondary 130 mm Cogan Case series Belgium Secondary <135 mm Donaldson Case-controlled clinical trial 17 Fenske Prospective observational study 18 Fichman Case series and rechallenge study Australia Secondary <134 mm 3 80 Germany Secondary <130 mm 5 81 USA Secondary 118 mm Fourlanos Case reports Australia (Not specified) (Not specified) Frenkel Controlled clinical trial The Netherlands (Not specified) (Not specified) Friedman Prospective controlled study Israel Secondary <130 mm Ghose Case reports UK Secondary (Not specified) Hajjar Case reports USA Secondary 130 mm Hoorn Prospective cohort The Netherlands Secondary 125 mm 7 88 study 25 Hung Case series China Secondary 113 mm Hwang Case series Korea Secondary 128 mm Johnston Case series UK Secondary 130 mm Jolobe Case series (letter) UK (Not specified) <120 mm Kalksma Case reports The Netherlands Primary <120 mm Kinoshita Case series Japan National registry (Not specified)

5 (Location not specified) 31 Mackay Case series New Zealand Secondary 132 mm Malin Case series USA Secondary <130 mm Mathew Case reports Australia (Not specified) 130 mm McDowell Retrospective cohort UK Primary 130 mm 6 98 study 35 Mouallem Case reports Israel Secondary 117 mm Musch Consecutive case Belgium Secondary 130 mm series 37 Oles Case reports USA Primary 129 mm Pinnock Case series Channel Islands Secondary 128 mm Rask Case reports Sweden Primary 121 mm Rastogi Retrospective, casecontrolled study 41 Rodenburg Population-based cohort study USA Secondary <135 mm Netherlands Primary 135 mm Rosner Case reports USA Secondary 122 mm Shapiro Prospective, Israel Secondary 125 mm observational, noninterventional study 44 Sharabi Case series USA Secondary <135 mm Sonnenblick Case series Israel Clinical trial <120 mm Sonnenblick Case reports Israel Secondary <115 mm Takeshita Case reports Japan Secondary 118 mm Tarssanen Case reports Finland Secondary 111 mm Thuesen Case reports Denmark Primary 113 mm van Case-controlled clinical The Netherlands Secondary (Not specified) Brummelen trial 51 Van Case reports The Netherlands Primary <135 mm Wijngaarden 52 Yong Case series Australia Secondary <135mM

6 53 Zalin Case reports UK Secondary 129 mm

7 Table S3 Meta-analyses of demographic characteristics of patients with Thiazide-Induced Hyponatremia by analysis of quality score, year of publication and age of patient. Symptoms Quality Score high Quality Score Low Year of publication Later Year of publication earlier Age of population younger Age of population older Mean 95% CI I 2 (%) Mean 95% CI I 2 (%) Mean 95% CI I 2 (%) Mean 95% CI I 2 (%) Mean 95% CI I 2 (%) Mean 95% CI I 2 (%) Gender (Female)* to to to to to to Age (years) to to to to to to Time to TIH (days) to to to to to to Prevalence estimates from meta-analysis and confidence intervals are all expressed as proportions. 7

8 Table S4: Meta-analyses of clinical characteristics of patients with Thiazide-Induced Hyponatremia by analysis of quality score, year of publication and age of patient. Prevalence estimates from meta-analysis and confidence intervals are all expressed as proportions. Symptoms Quality Score high Quality Score Low Year of publication Later Year of publication earlier Age of population younger Age of population older Prop 95% CI I 2 (%) Prop 95% CI I 2 (%) Prop 95% CI I 2 (%) Prop 95% CI I 2 (%) Prop 95% CI I 2 (%) Prop 95% CI I 2 (%) Fatigue to to 0.65 * to to to to 0.55 * Dizziness to study - * to to 1.00 * to to Confusion to to to to to to Vomiting to to to to to to Falls to study - * to to 1.00 * to 1.00 * to Nausea to to to to to to Unconsciousness to to to to to to Weakness to to to to to * Neurological to to to 0.99 * to to to Symptoms Seizures to to to to to to Diabetes to to 0.78 * to to 0.41 * to to CVD to to to to to to CardioVascular Disease (CVD), Prop (proportion), Confidence Interval (CI). None of these factors explain the high levels of heterogeneity between these studies. Prevalence estimates from meta-analysis and Confidence intervals are all expressed as proportions, * denotes less than 3 studies and therefore not possible to assess heterogeneity 8

9 Table S5: Meta-analyses of drug history of patients with Thiazide-Induced Hyponatremia by analysis of quality score, year of publication and age of patient. Prevalence estimates from meta-analysis and confidence intervals are all expressed as proportions. Symptoms Quality Score high Quality Score Low Year of publication Later Year of publication earlier Age of population younger Age of population older Proportion 95% CI I 2 (%) Proportion 95% CI I 2 (%) Proportion 95% CI I 2 (%) Proportion 95% CI I 2 (%) Proportion 95% CI I 2 (%) Proportion 95% CI I 2 (%) Thiazide HCTZ to to to to to to Indapamide to to to * 1 study - * to Moduretic to to to to to to Bendroflumethiazide to to 0.54 * to to to study - * Chlortalidone to to 0.78 * to to 0.40 * to to 0.79 * Other Drugs ACE inhibitor to to 1.00 * to study - * to 0.99 * to NSAID to to 0.71 * to to 0.71 * to to Non-thiazide diuretics to 0.22 * to to study - * to 0.98 * to Antidepressants to to 0.82 * to * to 1.00 * to HydroChloroThiaZide (HCTZ), AngioTensin Converting enzyme (ACE) inhibitor, Non-Steroidal AntiInflammatory Drug (NSAID), Prop (proportion), Confidence Interval (CI).None of these factors explain the high levels of heterogeneity between these studies. Prevalence estimates from meta-analysis and Confidence intervals are all expressed as proportions, * denotes less than 3 studies and therefore not possible to assess heterogeneity 9

10 Table S6: Meta-analyses of laboratory characteristics of patients with Thiazide-Induced Hyponatremia by analysis of quality score, year of publication and age of patient. Symptoms Quality Score high Quality Score Low Year of publication Later Year of publication earlier Age of population younger Age of population older Mean 95% CI I 2 (%) Mean 95% CI I 2 (%) Mean 95% CI I 2 (%) Mean 95% CI I 2 (%) Mean 95% CI I 2 (%) Mean 95% CI I 2 (%) Serum sodium (mm) to to to to to to Serum potassium to to to to to to (mm) Serum creatinine to to to to to to (µmol/l) Serum osmolality to to to to to to mosmkg Urine osmolality to to to to to to mosmkg Urine sodium to to to to to to (mm) Urine sodium 116.5mm to Prevalence estimates from meta-analysis and confidence intervals are all expressed as proportions. None of these factors explain the high levels of heterogeneity between these studies. Prevalence estimates from meta-analysis and Confidence intervals are all expressed as proportions. 10

11 Meta-analyses graphs Clinical characteristics and symptoms Figure S1 Proportion of patients with thiazide-induced hyponatremia who were female. 11

12 Aaseth (0.292, 1.000) Ambrosi (0.292, 1.000) Ashraf (0.290, 0.963) Bain (0.158, 1.000) Bayer (0.364, 0.793) Booker (0.541, 1.000) Canning (0.398, 1.000) Chapman (0.755, 0.883) Chow (0.635, 0.759) Clayton (0.299, 0.925) Coenraad (0.094, 0.992) Cogan (0.398, 1.000) Fenske (0.518, 0.997) Fichman (0.593, 0.932) Frenkel (0.316, 0.861) Friedman (0.390, 0.940) Hajjar (0.158, 1.000) Hung (0.099, 0.816) Hwang (0.768, 1.000) Johnston (0.292, 1.000) Jolobe (0.651, 0.956) Kalksma (0.094, 0.992) Kinoshita (0.458, 0.773) Malin (0.349, 0.968) Mathew (0.628, 0.861) McDowell (0.688, 0.784) Mouallem (0.006, 0.806) Musch (0.384, 0.837) Oles (0.292, 1.000) Pinnock (0.398, 1.000) Rask (0.292, 1.000) Rastogi (0.686, 0.728) Rodenburg (0.807, 0.916) Rosner (0.158, 1.000) Shapiro (0.830, 0.975) Sharabi (0.765, 0.880) Takeshita (0.158, 1.000) Tarssanen (0.292, 1.000) Thuesen (0.158, 1.000) Van Wijngaarden (0.158, 1.000) Yong (0.715, 1.000) Zalin (0.245, 0.915) van Brummelen (0.000, 0.842) combined (0.744, 0.823)

13 Figure S2. Summary of mean age in patients with thiazide-induced hyponatremia. Authorship Year of publication ES (95% CI) % Weight Aaseth 2001 Ambrosi 2004 Ashraf 1981 Booker 1984 Canning 1988 Chow 2004 Clayton 2006 Coenraad 2003 Cogan Fenske Fichman 1971 Friedman Hajjar Hung 2002 Hwang 2010 Johnston 1989 Kalksma 2002 Kinoshita 2011 Mackay 1983 McDowell 2010 Mouallem 1983 Musch 2001 Oles 1984 Pinnock 1978 Rask 1996 Rastogi 2012 Rosner 2004 Shapiro 2010 Sharabi 2002 Sonnenblick 1986 Takeshita 2010 Tarssanen 1980 Thuesen 1980 Van Wijngaarden 2010 Yong 2011 Zalin 1984 Overall (I squared = 92.6%, p = 0.000) NOTE: Weights are from random effects analysis Age (years) (62.55, 87.45) (80.36, 88.30) (60.75, 75.83) (61.72, 72.28) (66.57, 83.43) (74.82, 77.18) (55.03, 73.37) (65.22, 96.12) (74.34, 78.66) (68.16, 83.84) (57.13, 64.79) (65.60, 76.40) (87.56, 93.44) (69.69, 83.73) (72.86, 79.14) (62.40, 68.94) (79.21, 92.79) (73.83, 78.97) (63.54, 74.64) (71.97, 74.23) (52.21, 81.29) (70.50, 79.50) (70.82, 91.84) (70.57, 84.43) (72.14, 86.52) (70.53, 71.67) (52.93, 94.07) (79.99, 84.21) (75.00, 77.80) (70.56, 84.04) (57.28, 84.72) (62.24, 85.76) (49.25, 72.75) (73.69, 91.31) (78.27, 85.13) (57.98, 72.26) (73.03, 76.85)

14 Figure S3. Summary of mean Body Mass Index in patients with thiazideinduced hyponatremia (Kg/m 2 ). Year of % Authorship publication ES (95% CI) Weight Chow (21.91, 22.89) Rastogi (27.10, 27.70) Overall (I squared = 99.7%, p = 0.000) (20.00, 29.80) NOTE: Weights are from random effects analysis BMI (kg/m2) 14

15 Figure S4. Summary of mean duration from thiazide initiation to presentation with thiazide-induced hyponatremia (days). Year of % Authorship publication ES (95% CI) Weight Ashraf Bain Booker 1984 Clayton 2006 Fichman 1971 Friedman 1989 Hung 2002 Kalksma 2002 McDowell 2010 Mouallem Oles Rask 1996 Sonnenblick 1989 Takeshita 2010 Zalin 1984 Overall (I squared = 96.9%, p = 0.000) 5.67 (1.82, 9.52) 4.00 (2.05, 5.95) 7.60 (3.20, 12.00) (145.30, ) (2.06, 32.80) 6.67 (3.46, 9.88) ( , ) (511.10, ) (59.33, 70.43) ( 48.50, ) 8.33 ( 2.18, 18.84) (1.16, 30.84) 9.50 (3.27, 15.73) (15.60, 74.40) (59.71, ) (7.97, 30.09) NOTE: Weights are from random effects analysis Time To TIH (days) 15

16 Figure S5. Proportion of patients with thiazide-induced hyponatremia who reported fatigue. Aaseth (0.094, 0.992) Ashraf (0.004, 0.579) Canning (0.398, 1.000) Chapman (0.020, 0.133) Chow (0.421, 0.556) Hajjar (0.013, 0.987) Rosner (0.158, 1.000) Zalin (0.032, 0.651) combined (0.213, 0.719) Figure S6. Proportion of patients with thiazide-induced hyponatremia who reported dizziness. Aaseth (0.008, 0.906) Chow (0.461, 0.596) Friedman (0.023, 0.518) Hung (0.184, 0.901) Mathew (0.061, 0.254) Rask (0.008, 0.906) Sharabi (0.101, 0.211) Tarssanen (0.008, 0.906) combined (0.148, 0.508) 16

17 Figure S7. Proportion of patients with thiazide-induced hyponatremia who reported confusion. Aaseth (0.008, 0.906) Ambrosi (0.094, 0.992) Ashraf (0.037, 0.710) Bain (0.013, 0.987) Bayer (0.407, 0.828) Booker (0.359, 0.996) Chapman (0.085, 0.250) Chow (0.127, 0.231) Friedman (0.234, 0.833) Hung (0.004, 0.579) Johnston (0.008, 0.906) Kalksma (0.008, 0.906) Mathew (0.139, 0.372) Oles (0.292, 1.000) Pinnock (0.398, 1.000) Rask (0.094, 0.992) Rosner (0.013, 0.987) Shapiro (0.582, 0.814) Sharabi (0.115, 0.229) Van Wijngaarden (0.013, 0.987) Yong (0.234, 0.833) Zalin (0.032, 0.651) combined (0.328, 0.561) 17

18 Figure S8. Proportion of patients with thiazide-induced hyponatremia who reported vomiting. Aaseth (0.094, 0.992) Ambrosi (0.094, 0.992) Ashraf (0.184, 0.901) Booker (0.223, 0.957) Chapman (0.067, 0.222) Chow (0.283, 0.412) Friedman (0.002, 0.413) Hung (0.290, 0.963) Johnston (0.008, 0.906) Sharabi (0.307, 0.453) Tarssanen (0.008, 0.906) Yong (0.060, 0.610) Zalin (0.003, 0.527) combined (0.253, 0.449) Figure S9. Proportion of patients with thiazide-induced hyponatremia who reported falls. Ambrosi (0.09, 0.99) Canning (0.40, 1.00) Chow (0.12, 0.22) Friedman (0.17, 0.77) Yong (0.11, 0.69) combined 0.48 (0.20, 0.77) 18

19 Figure S10. Proportion of patients with thiazide-induced hyponatremia who reported nausea. Aaseth (0.094, 0.992) Ashraf (0.184, 0.901) Bayer (0.407, 0.828) Booker (0.223, 0.957) Chapman (0.104, 0.277) Friedman (0.023, 0.518) Hung (0.099, 0.816) Johnston (0.292, 1.000) Mathew (0.019, 0.167) Rask (0.008, 0.906) Sharabi (0.204, 0.338) Van Wijngaarden (0.013, 0.987) Yong (0.060, 0.610) Zalin (0.003, 0.527) combined (0.244, 0.481) 19

20 Figure S11. Proportion of patients with thiazide-induced hyponatremia who were reported to be unconscious. Aaseth (0.008, 0.906) Ashraf (0.004, 0.579) Bain (0.013, 0.987) Booker (0.223, 0.957) Chapman (0.013, 0.117) Friedman (0.060, 0.610) Kalksma (0.008, 0.906) Mathew (0.004, 0.119) Rask (0.008, 0.906) Takeshita (0.158, 1.000) Van Wijngaarden (0.013, 0.987) combined (0.148, 0.476) 20

21 Figure S12. Proportion of patients with thiazide-induced hyponatremia who reported weakness. Aaseth (0.008, 0.906) Ambrosi (0.292, 1.000) Bain (0.013, 0.987) Bayer (0.407, 0.828) Booker (0.004, 0.641) Ghose (0.158, 1.000) Hung (0.004, 0.579) Kalksma (0.008, 0.906) Oles (0.008, 0.906) Pinnock (0.006, 0.806) Rosner (0.013, 0.987) Sharabi (0.323, 0.470) Van Wijngaarden (0.158, 1.000) Zalin (0.003, 0.527) combined (0.321, 0.579) 21

22 Figure S13. Proportion of patients with thiazide-induced hyponatremia who reported neurological symptoms. Aaseth (0.008, 0.906) Ashraf (0.004, 0.579) Bain (0.013, 0.987) Booker (0.004, 0.641) Oles (0.008, 0.906) Tarssanen (0.008, 0.906) Van Wijngaarden (0.158, 1.000) Yong (0.715, 1.000) combined (0.219, 0.797) 22

23 Figure S14. Proportion of patients with thiazide-induced hyponatremia who had seizures. Ashraf (0.037, 0.710) Booker (0.004, 0.641) Chapman (0.003, 0.083) Chow (0.001, 0.032) Johnston (0.292, 1.000) Mathew (0.039, 0.212) Oles (0.008, 0.906) Rosner (0.013, 0.987) Yong (0.023, 0.518) Zalin (0.032, 0.651) combined (0.083, 0.332) Figure S15. Proportion of patients with thiazide-induced hyponatremia who had cardiovascular disease. Aaseth (0.094, 0.992) Ashraf (0.025, 1.000) Bain (0.158, 1.000) Booker (0.004, 0.641) Friedman (0.023, 0.518) Hajjar (0.013, 0.987) Kalksma (0.292, 1.000) Mouallem (0.398, 1.000) Pinnock (0.068, 0.932) Shapiro (0.239, 0.482) Sharabi (0.164, 0.290) Tarssanen (0.008, 0.906) combined (0.330, 0.647) 23

24 Figure S16. Proportion of patients with thiazide-induced hyponatremia who had diabetes mellitus. Al Qahtani (0.680, 0.763) Chow (0.192, 0.309) Friedman (0.002, 0.413) Kalksma (0.008, 0.906) McDowell (0.023, 0.069) Rastogi (0.219, 0.258) Sharabi (0.144, 0.266) Takeshita (0.013, 0.987) Tarssanen (0.008, 0.906) combined (0.120, 0.475) 24

25 DRUGS Figure S17. Proportion of patients with thiazide-induced hyponatremia who took hydrochlorothiazide. Aaseth (0.008, 0.906) Al Qahtani (0.455, 0.547) Ambrosi (0.008, 0.906) Ashraf (0.290, 0.963) Chapman (0.002, 0.048) Fenske (0.664, 1.000) Fichman (0.425, 0.820) Hajjar (0.013, 0.987) Kalksma (0.094, 0.992) Kinoshita (0.912, 1.000) Liamis (0.932, 1.000) Oles (0.292, 1.000) Pinnock (0.006, 0.806) Rastogi (0.607, 0.652) Rosner (0.158, 1.000) Sonnenblick (0.398, 1.000) Takeshita (0.013, 0.987) Van Wijngaarden (0.013, 0.987) van Brummelen (0.158, 1.000) combined (0.519, 0.826) 25

26 Figure S18. Proportion of patients with thiazide-induced hyponatremia who took indapamide. Al Qahtani (0.4527, ) Ambrosi (0.0943, ) Chapman (0.4802, ) Fourlanos (0.5200, ) Hung (0.2904, ) Jolobe (0.0010, ) McDowell (0.0123, ) Yong (0.7151, ) combined (0.2330, ) 26

27 Figure S19. Proportion of patients with thiazide-induced hyponatremia who took Moduretic (hydrochlorothiazide with amiloride). Aaseth (0.008, 0.906) Adams (0.013, 0.987) Bain (0.013, 0.987) Bayer (0.282, 0.718) Canning (0.398, 1.000) Chapman (0.342, 0.506) Donaldson (0.398, 1.000) Fourlanos (0.318, 0.428) Friedman (0.715, 1.000) Hung (0.037, 0.710) Johnston (0.292, 1.000) Jolobe (0.202, 0.594) Mackay (0.390, 0.940) Mathew (0.938, 1.000) Pinnock (0.194, 0.994) Rask (0.292, 1.000) Sonnenblick (0.025, 1.000) Tarssanen (0.292, 1.000) Zalin (0.631, 1.000) combined (0.572, 0.866) 27

28 Figure S20. Proportion of patients with thiazide-induced hyponatremia who took bendroflumethiazide (bendrofluazide). Aaseth (0.008, 0.906) Bain (0.013, 0.987) Bayer (0.107, 0.502) Booker (0.118, 0.882) Clayton (0.664, 1.000) Fichman (0.001, 0.204) Jolobe (0.172, 0.557) McDowell (0.942, 0.984) combined (0.152, 0.880) Figure S21. Proportion of patients with thiazide-induced hyponatremia who took Dyazide (hydrochlorothiazide with triamterene). Bayer (0.078, 0.454) Jolobe (0.009, 0.251) Mackay (0.060, 0.610) combined (0.078, 0.321)

29 Figure S22. Proportion of patients with thiazide-induced hyponatremia who took chlortalidone. Booker (0.0042, ) Fichman (0.0936, ) Kalksma (0.0084, ) McDowell (0.0007, ) Rastogi (0.0077, ) Van Wijngaarden (0.0126, ) combined (0.0197, ) Figure S23. Proportion of patients with thiazide-induced hyponatremia who were also taking an Angiotensin Converting Enzyme (ACE) inhibitor. Kalksma (0.09, 0.99) Mouallem (0.40, 1.00) Rastogi (0.57, 0.62) Sharabi (0.18, 0.31) Yong (0.06, 0.61) combined 0.51 (0.27, 0.75) 29

30 Figure S24. Proportion of patients with thiazide-induced hyponatremia who were also taking non-steroidal anti-inflammatory drugs. Aaseth (0.008, 0.906) Chow (0.108, 0.206) Kalksma (0.008, 0.906) Oles (0.008, 0.906) Rastogi (0.317, 0.361) Rosner (0.158, 1.000) combined (0.178, 0.492) Figure S25. Proportion of patients with thiazide-induced hyponatremia who were also prescribed a non-thiazide diuretic. Ambrosi (0.008, 0.906) Coenraad (0.008, 0.906) Kalksma (0.292, 1.000) Mouallem (0.398, 1.000) Rastogi (0.183, 0.221) combined (0.190, 0.913) 30

31 Figure S26. Proportion of patients with thiazide-induced hyponatremia who were also taking antidepressants. Ambrosi (0.094, 0.992) Hung (0.004, 0.579) Kalksma (0.008, 0.906) Rastogi (0.189, 0.227) Rosner (0.158, 1.000) Shapiro (0.212, 0.451) combined (0.191, 0.469) Figure S27. Proportion of patients with thiazide-induced hyponatremia who were also taking potassium supplements. Rastogi (0.144, 0.179) Tarssanen (0.008, 0.906) combined (0.145, 0.179) 31

32 Figure S28. Proportion of patients with thiazide-induced hyponatremia who were also taking angiotensin II receptor blockers. Kinoshita (0.912, 1.000) Rastogi (0.097, 0.127) Takeshita (0.013, 0.987) combined (0.001, 0.955) 32

33 Laboratory characteristics Figure S29. Summary of mean concentration of serum sodium in patients with thiazide-induced hyponatremia (mm). Authorship Year of publication ES (95% CI) % Weight Aaseth 2001 Ambrosi 2004 Ashraf 1981 Bain Bayer Booker 1984 Canning 1988 Chow (Female) Chow (Male) Clayton 2006 Coenraad 2003 Donaldson Fenske Fichman 1971 Friedman 1989 Hajjar Hung Hwang 2010 Johnston 1989 Kalksma Kinoshita Mackay Mathew 1990 McDowell 2010 Musch 2001 Oles 1984 Pinnock 1978 Rask 1996 Rosner 2004 Sharabi 2002 Sonnenblick 1986 Sonnenblick 1989 Takeshita 2010 Tarssanen 1980 Thuesen 1980 Van Wijngaarden 2010 Yong 2011 Zalin 1984 Overall (I squared = 99.1%, p = 0.000) NOTE: Weights are from random effects analysis (109.34, ) (110.72, ) (99.81, ) (110.06, ) (124.78, ) (105.39, ) (107.20, ) (112.75, ) (115.07, ) (116.09, ) (117.05, ) (126.90, ) (111.20, ) (106.08, ) (116.35, ) (127.05, ) (105.74, ) (114.86, ) (88.88, ) (106.93, ) (112.30, ) (124.24, ) (116.35, ) (127.79, ) (123.75, ) (109.92, ) (111.21, ) (113.55, ) (102.26, ) (119.12, ) (107.47, ) (102.64, ) (109.17, ) (104.46, ) (108.56, ) (108.05, ) (107.41, ) (112.71, ) (113.40, ) Serum sodium concentration (mm) 33

34 Figure S30. Summary of mean concentration of serum potassium in patients with thiazide-induced hyponatremia (mm). Year of % Authorship publication ES (95% CI) Weight Aaseth 2001 Booker 1984 Canning 1988 Chow 2004 Clayton 2006 Fenske 2009 Fichman 1971 Hung 2002 Hwang 2010 Johnston 1989 Kalksma 2002 McDowell 2010 Musch 2001 Oles 1984 Pinnock 1978 Rask 1996 Rosner 2004 Sharabi 2002 Sonnenblick 1986 Takeshita 2010 Tarssanen 1980 Thuesen 1980 Van Wijngaarden 2010 Yong 2011 van Brummelen 1978 Overall (I squared = 96.8%, p = 0.000) NOTE: Weights are from random effects analysis 2.97 (2.30, 3.64) 2.84 (2.37, 3.31) 3.83 (3.61, 4.05) 3.30 (3.20, 3.40) 3.69 (2.74, 4.64) 3.80 (3.28, 4.32) 2.88 (2.70, 3.06) 3.40 (2.53, 4.27) 3.50 (3.19, 3.81) 3.00 (2.15, 3.85) 2.73 (2.56, 2.90) 4.14 (4.07, 4.21) 3.50 (3.23, 3.77) 2.90 (2.32, 3.48) 2.95 (2.85, 3.05) 3.43 (2.81, 4.05) 3.30 (2.51, 4.09) 3.80 (3.70, 3.90) 3.70 (3.48, 3.92) 2.55 (2.26, 2.84) 3.35 (2.46, 4.24) 3.15 (2.07, 4.23) 3.20 (2.62, 3.78) 2.80 (2.45, 3.15) 2.85 (2.56, 3.14) 3.26 (3.02, 3.51) Serum potassium concentration (mm) 34

35 Figure S31. Summary of mean concentration of serum creatinine in patients with thiazide-induced hyponatremia (µmol/l). Year of % Authorship publication ES (95% CI) Weight Aaseth (68.21, ) 5.48 Canning (63.18, 74.32) 7.35 Chow (44.11, 49.89) 7.42 Fenske (45.62, ) 2.30 Fichman (64.18, 75.78) 7.34 Hwang (52.62, 71.14) 7.17 Johnston (48.48, 95.52) 5.92 Kalksma (91.23, ) 7.00 Mouallem (137.15, ) 7.36 Musch (58.90, 82.74) 6.99 Rask (59.60, 63.06) 7.44 Sharabi (84.33, ) 6.91 Takeshita (39.96, 57.28) 7.20 Thuesen (63.56, 69.44) 7.42 Yong (46.46, 77.54) 6.70 Overall (I squared = 98.8%, p = 0.000) (64.14, 89.43) NOTE: Weights are from random effects analysis Serum Creatinine (umol/l) 35

36 Figure S32. Summary of mean serum osmolality in patients with thiazideinduced hyponatremia (mosm/kg). Year of % Authorship publication ES (95% CI) Weight Aaseth (246.76, ) Booker (219.42, ) 7.03 Chow (235.64, ) Fichman (233.14, ) 9.24 Hung (226.84, ) 8.28 Hwang (235.24, ) Rask (243.94, ) 8.35 Sonnenblick (237.00, ) Sonnenblick (204.09, ) 2.10 Takeshita (199.65, ) 1.77 Yong (226.60, ) Overall (I squared = 80.3%, p = 0.000) (235.94, ) NOTE: Weights are from random effects analysis Serum Osmolality (mosm/kg) 36

37 Figure S33. Summary of mean urinary osmolality in patients with thiazideinduced hyponatremia (mosm/kg). Year of % Authorship publication ES (95% CI) Weight Booker (349.78, ) 6.69 Chow (371.13, ) Fenske (215.71, ) 8.50 Fichman (264.63, ) 8.00 Hung (281.60, ) 4.44 Hwang (328.67, ) 9.00 Johnston (178.19, ) 3.32 Musch (421.85, ) 9.13 Rask (114.91, ) 1.42 Rosner (377.80, ) Takeshita (458.64, ) Yong (359.19, ) Overall (I squared = 80.5%, p = 0.000) (370.31, ) NOTE: Weights are from random effects analysis Urine osmolality (mosm/kg) Figure S34. Summary of mean concentration of urinary sodium in patients with thiazide-induced hyponatremia (mm). Year of % Authorship publication ES (95% CI) Weight Aaseth (11.96, 21.38) Booker (63.91, ) 8.16 Fenske (43.09, 84.91) 9.53 Fichman (18.99, 55.37) 9.88 Hung (58.56, ) 6.45 Hwang (64.29, ) 8.74 Johnston (7.71, 78.29) 7.53 Musch (46.01, 81.99) 9.90 Rosner (40.60, ) 5.26 Takeshita ( 13.08, ) 2.53 Thuesen (45.59, 55.41) Yong (62.35, 97.45) 9.95 Overall (I squared = 93.9%, p = 0.000) (46.97, 81.00) NOTE: Weights are from random effects analysis Urinary sodium concentration (mm) 37

38 PRISMA statement for systematic reviews Section/topic* Item* No* Checklist*item* Reported*on* page*no* Title* Title* 1 Identifythereportasasystematicreview, meta?analysis,orboth 0 Abstract* Structured* summary* Introduction* 2 Provideastructuredsummaryincluding,as applicable,background,objectives,data sources,studyeligibilitycriteria, participants,interventions,studyappraisal andsynthesismethods,results,limitations, conclusionsandimplicationsofkeyfindings, systematicreviewregistrationnumber 2 Rationale* 3 Describetherationaleforthereviewinthe contextofwhatisalreadyknown Objectives* 4 Provideanexplicitstatementofquestions beingaddressedwithreferenceto participants,interventions,comparisons, outcomes,andstudydesign(picos) 4 4 Methods* Protocol*and* registration* Eligibility* criteria* Information* sources* 5 Indicateifareviewprotocolexists,ifand whereitcanbeaccessed(suchasweb address),and,ifavailable,provide registrationinformationincluding registrationnumber 6 Specifystudycharacteristics(suchasPICOS, lengthoffollow?up)andreport characteristics(suchasyearsconsidered, language,publicationstatus)usedascriteria foreligibility,givingrationale 7 Describeallinformationsources(suchas databaseswithdatesofcoverage,contact withstudyauthorstoidentifyadditional studies)inthesearchanddatelastsearched

39 Search* 8 Presentfullelectronicsearchstrategyforat leastonedatabase,includinganylimits used,suchthatitcouldberepeated Study*selection* 9 Statetheprocessforselectingstudies(that is,screening,eligibility,includedin systematicreview,and,ifapplicable, includedinthemeta?analysis) 5 5 Data*collection* process* 10 Describemethodofdataextractionfrom reports(suchaspilotedforms, independently,induplicate)andany processesforobtainingandconfirmingdata frominvestigators 5 Data*items* 11 Listanddefineallvariablesforwhichdata weresought(suchaspicos,funding sources)andanyassumptionsand simplificationsmade 5 Risk*of*bias*in* individual* studies* Summary* measures* Synthesis*of* results* Risk*of*bias* across*studies* Additional* analyses* Results* 12 Describemethodsusedforassessingriskof biasofindividualstudies(including specificationofwhetherthiswasdoneatthe studyoroutcomelevel),andhowthis informationistobeusedinanydata synthesis 13 Statetheprincipalsummarymeasures(such asriskratio,differenceinmeans). 14 Describethemethodsofhandlingdataand combiningresultsofstudies,ifdone, includingmeasuresofconsistency(suchas I 2 statistic)foreachmeta?analysis 15 Specifyanyassessmentofriskofbiasthat mayaffectthecumulativeevidence(suchas publicationbias,selectivereportingwithin studies) 16 Describemethodsofadditionalanalyses (suchassensitivityorsubgroupanalyses, meta?regression),ifdone,indicatingwhich werepre?specified 5?6 6 5?6 5?6 5?6 Study*selection* 17 Givenumbersofstudiesscreened,assessed foreligibility,andincludedinthereview, 6 39

40 Study* characteristics* Risk*of*bias* within*studies* Results*of* individual* studies* Synthesis*of* results* Risk*of*bias* across*studies* Additional* analysis* Discussion* Summary*of* evidence* withreasonsforexclusionsateachstage, ideallywithaflowdiagram 18 Foreachstudy,presentcharacteristicsfor whichdatawereextracted(suchasstudy size,picos,follow?upperiod)andprovide thecitations 19 Presentdataonriskofbiasofeachstudy and,ifavailable,anyoutcome?level assessment(seeitem12). 20 Foralloutcomesconsidered(benefitsor harms),presentforeachstudy(a)simple summarydataforeachinterventiongroup and(b)effectestimatesandconfidence intervals,ideallywithaforestplot 21 Presentresultsofeachmeta?analysisdone, includingconfidenceintervalsandmeasures ofconsistency 22 Presentresultsofanyassessmentofriskof biasacrossstudies(seeitem15) 23 Giveresultsofadditionalanalyses,ifdone (suchassensitivityorsubgroupanalyses, meta?regression)(seeitem16) 24 Summarisethemainfindingsincludingthe strengthofevidenceforeachmainoutcome; considertheirrelevancetokeygroups(such ashealthproviders,users,andpolicy makers) 6?9and online supplement 6?9and online supplement 6?9and online supplement 6?9and online supplement 6?9and online supplement 6?9and online supplement 10?11 Limitations* 25 Discusslimitationsatstudyandoutcome level(suchasriskofbias),andatreview level(suchasincompleteretrievalof identifiedresearch,reportingbias) Conclusions* 26 Provideageneralinterpretationofthe resultsinthecontextofotherevidence,and implicationsforfutureresearch 12?13 14 Funding* 40

41 Funding* 27 Describesourcesoffundingforthe systematicreviewandothersupport(such assupplyofdata)androleoffundersforthe systematicreview 14 41