Objective

Determine whether crushing tablet medication creates an aerosol exposure risk for healthcare workers, and whether the risk differs by pill crusher and crushing method.

Background

The three pill crushers used included MAXCRUSH, Silent Knight, and SafeCrush. Two different pill containers were used with the MAXCRUSH pill crusher labeled as “cup” and “packaged”.

Tablet medications were crushed in a controlled laboratory setting. Medication preparation was divided into three phases. “Crushing” was when the pill was actively being crushed. “Transitioning” included the time immediately following crushing until the medication was poured from the container. “Pouring” was when the crushed tablet was poured from the container into a cup of water.

Breathing zone measurements were taken under controlled and uncontrolled conditions. “Controlled” involved carefully crushing and pouring the crushed medication out of the pill container and “uncontrolled” involved vigorously crushing and pouring the crushed medication out of the pill container. Twenty trials were performed for each condition.

Normality Visuzlization

QQ Plot

Normality Visualization Log Transformed

QQ Plot Log Transformed

Boxplot Comparing Containers in Controlled and Uncontrolled Scenarios

Shapiro-Wilk Tests

Controlled Scenario

## # A tibble: 4 × 2
##   container     p_value
##   <chr>           <dbl>
## 1 cup             0.263
## 2 packaged        0.651
## 3 safecrush       0.215
## 4 silent knight   0.578

Indicates data follows a normal distribution

Uncontrolled Scenario

## # A tibble: 4 × 2
##   container     p_value
##   <chr>           <dbl>
## 1 cup            0.646 
## 2 packaged       0.0518
## 3 safecrush      0.369 
## 4 silent knight  0.0872

Indicates data follows a normal distribution

Bartlett Test

Controlled Scenario

## 
##  Bartlett test of homogeneity of variances
## 
## data:  log_particle_count by container
## Bartlett's K-squared = 1.597, df = 3, p-value = 0.6601

Indicates that variance of groups are equal

Uncontrolled Scenario

## 
##  Bartlett test of homogeneity of variances
## 
## data:  log_particle_count by container
## Bartlett's K-squared = 1.1122, df = 3, p-value = 0.7741

Indicates that variance of groups are equal

Levenes Tests

Controlled Scenario

## Levene's Test for Homogeneity of Variance (center = median)
##       Df F value Pr(>F)
## group  3  0.3387 0.7974
##       76

Indicates that variance of groups are equal

Uncontrolled Scenario

## Levene's Test for Homogeneity of Variance (center = median)
##       Df F value Pr(>F)
## group  3  0.3425 0.7947
##       75

Indicates that variance of groups are equal

ANOVA

Controlled Scenario

##             Df Sum Sq Mean Sq F value  Pr(>F)    
## container    3  33.13  11.044   13.34 4.4e-07 ***
## Residuals   76  62.91   0.828                    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Indicates that at least one group is significantly different from the others

Tukeys HSD Test

Controlled Scenario

##   Tukey multiple comparisons of means
##     95% family-wise confidence level
## 
## Fit: aov(formula = log_particle_count ~ container, data = .)
## 
## $container
##                               diff         lwr        upr     p adj
## packaged-cup             1.6574293  0.90168833  2.4131702 0.0000010
## safecrush-cup            0.1833568 -0.57238415  0.9390977 0.9196153
## silent knight-cup        0.6988412 -0.05689971  1.4545821 0.0801888
## safecrush-packaged      -1.4740725 -2.22981340 -0.7183316 0.0000130
## silent knight-packaged  -0.9585880 -1.71432896 -0.2028471 0.0071538
## silent knight-safecrush  0.5154844 -0.24025649  1.2712254 0.2852687

Indicates significant difference for packaged-cup, safecrush-packaged, and silent knight-packaged

ANOVA

Uncontrolled Scenario

##             Df Sum Sq Mean Sq F value  Pr(>F)    
## container    3 112.43   37.48   41.86 5.3e-16 ***
## Residuals   75  67.14    0.90                    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Indicates that at least one group is significantly different from the others

Tukeys HSD Test

Uncontrolled Scenario

##   Tukey multiple comparisons of means
##     95% family-wise confidence level
## 
## Fit: aov(formula = log_particle_count ~ container, data = .)
## 
## $container
##                               diff        lwr         upr     p adj
## packaged-cup             3.0651508  2.2789751  3.85132647 0.0000000
## safecrush-cup            0.9719537  0.1857780  1.75812940 0.0092200
## silent knight-cup        2.3406182  1.5441653  3.13707115 0.0000000
## safecrush-packaged      -2.0931971 -2.8793728 -1.30702137 0.0000000
## silent knight-packaged  -0.7245326 -1.5209855  0.07192037 0.0877120
## silent knight-safecrush  1.3686645  0.5722116  2.16511744 0.0001336

Indicates significant difference for packaged-cup, safecrush-cup, silent knight-cup, safecrush-packaged, and silent knight-safecrush

Comparing Controlled & Uncontrolled for Each Container

Shapiro-Wilk Test MAXCRUSH with Pill Cup

## # A tibble: 2 × 2
##   scenario     p_value
##   <chr>          <dbl>
## 1 controlled     0.263
## 2 uncontrolled   0.646

Indicates data follows a normal distribution

Bartlett Test MAXCRUSH with Pill Cup

## 
##  Bartlett test of homogeneity of variances
## 
## data:  log_particle_count by scenario
## Bartlett's K-squared = 9.3168e-06, df = 1, p-value = 0.9976

Indicates that variance of groups are equal

Levenes Test MAXCRUSH with Pill Cup

## Levene's Test for Homogeneity of Variance (center = median)
##       Df F value Pr(>F)
## group  1  0.0103 0.9198
##       38

Indicates that variance of groups are equal

T Test MAXCRUSH with Pill Cup

## 
##  Two Sample t-test
## 
## data:  . and cup %>% filter(scenario == "uncontrolled") %>% pull(log_particle_count)
## t = -1.4811, df = 38, p-value = 0.1468
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -0.9009372  0.1396240
## sample estimates:
## mean of x mean of y 
##  5.183078  5.563735

Indicates no significant difference between groups

Shapiro-Wilk Test MAXCRUSH with Unit Dose Packaging

## # A tibble: 2 × 2
##   scenario     p_value
##   <chr>          <dbl>
## 1 controlled    0.651 
## 2 uncontrolled  0.0518

Indicates data follows a normal distribution

Bartlett Test MAXCRUSH with Unit Dose Packaging

## 
##  Bartlett test of homogeneity of variances
## 
## data:  log_particle_count by scenario
## Bartlett's K-squared = 0.58253, df = 1, p-value = 0.4453

Indicates that variance of groups are equal

Levenes Test MAXCRUSH with Unit Dose Packaging

## Levene's Test for Homogeneity of Variance (center = median)
##       Df F value Pr(>F)
## group  1  0.4265 0.5176
##       38

Indicates that variance of groups are equal

T Test MAXCRUSH with Unit Dose Packaging

## 
##  Two Sample t-test
## 
## data:  . and packaged %>% filter(scenario == "uncontrolled") %>% pull(log_particle_count)
## t = -6.1104, df = 38, p-value = 4.016e-07
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -2.380868 -1.195888
## sample estimates:
## mean of x mean of y 
##  6.840507  8.628885

Indicates significant difference between groups

Shapiro-Wilk Test Silent Knight

## # A tibble: 2 × 2
##   scenario     p_value
##   <chr>          <dbl>
## 1 controlled    0.578 
## 2 uncontrolled  0.0872

Indicates data follows a normal distribution

Bartlett Test Silent Knight

## 
##  Bartlett test of homogeneity of variances
## 
## data:  log_particle_count by scenario
## Bartlett's K-squared = 0.027252, df = 1, p-value = 0.8689

Indicates that variance of groups are equal

Levenes Test Silent Knight

## Levene's Test for Homogeneity of Variance (center = median)
##       Df F value Pr(>F)
## group  1   0.025 0.8752
##       37

Indicates that variance of groups are equal

T Test Silent Knight

## 
##  Two Sample t-test
## 
## data:  . and silent_knight %>% filter(scenario == "uncontrolled") %>% pull(log_particle_count)
## t = -6.08, df = 37, p-value = 4.893e-07
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -2.696417 -1.348450
## sample estimates:
## mean of x mean of y 
##  5.881919  7.904353

Indicates significant difference between groups

Shapiro-Wilk Test Safe Crush

## # A tibble: 2 × 2
##   scenario     p_value
##   <chr>          <dbl>
## 1 controlled     0.215
## 2 uncontrolled   0.369

Indicates data follows a normal distribution

Bartlett Test Test Safe Crush

## 
##  Bartlett test of homogeneity of variances
## 
## data:  log_particle_count by scenario
## Bartlett's K-squared = 0.023154, df = 1, p-value = 0.8791

Indicates that variance of groups are equal

Levenes Test Safe Crush

## Levene's Test for Homogeneity of Variance (center = median)
##       Df F value Pr(>F)
## group  1  0.0876 0.7689
##       38

Indicates that variance of groups are equal

T Test Safe Crush

## 
##  Two Sample t-test
## 
## data:  . and safe_crush %>% filter(scenario == "uncontrolled") %>% pull(log_particle_count)
## t = -3.9975, df = 38, p-value = 0.0002845
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -1.7613738 -0.5771331
## sample estimates:
## mean of x mean of y 
##  5.366435  6.535688

Indicates significant difference between groups

Phase Comparison for MAXCRUSH with Pill Cup

Shapiro-Wilk Test

Controlled Scenario

## # A tibble: 3 × 2
##   phase         p_value
##   <chr>           <dbl>
## 1 crushing        0.134
## 2 pouring         0.405
## 3 transitioning   0.828

Indicates data follows a normal distribution

Shapiro-Wilk Test

Uncontrolled Scenario

## # A tibble: 3 × 2
##   phase         p_value
##   <chr>           <dbl>
## 1 crushing        0.388
## 2 pouring         0.417
## 3 transitioning   0.421

Indicates data follows a normal distribution

Bartlett Test

Controlled Scenario

## 
##  Bartlett test of homogeneity of variances
## 
## data:  log_particle_count by phase
## Bartlett's K-squared = 15.142, df = 2, p-value = 0.0005152

Indicates that variance of groups are unequal

Bartlett Test

Uncontrolled Scenario

## 
##  Bartlett test of homogeneity of variances
## 
## data:  log_particle_count by phase
## Bartlett's K-squared = 1.9155, df = 2, p-value = 0.3837

Indicates that variance of groups are equal

Levenes Test

Controlled Scenario

## Levene's Test for Homogeneity of Variance (center = median)
##       Df F value   Pr(>F)   
## group  2  6.5417 0.002769 **
##       57                    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Indicates that variance of groups are unequal

Levenes Test

Uncontrolled Scenario

## Levene's Test for Homogeneity of Variance (center = median)
##       Df F value Pr(>F)
## group  2   1.559 0.2192
##       57

Indicates that variance of groups are equal

Kruskal Wallis Test

Controlled Scenario

## [1] 3.35799e-05

Indicates that at least one group is significantly different from the others

Dunns Test

Controlled Scenario

##                 Comparison          Z      P.unadj        P.adj
## 1       crushing - pouring -3.3464118 8.186472e-04 2.455942e-03
## 2 crushing - transitioning  0.9826405 3.257844e-01 9.773532e-01
## 3  pouring - transitioning  4.3290523 1.497524e-05 4.492571e-05

Indicates significant difference for crushing-pouring and pouring-transitioning

ANOVA Test

Uncontrolled Scenario

##             Df Sum Sq Mean Sq F value   Pr(>F)    
## phase        2  25.08  12.542   8.759 0.000482 ***
## Residuals   57  81.62   1.432                     
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Indicates that at least one group is significantly different from the others

Tukeys HSD Test

Uncontrolled Scenario

##   Tukey multiple comparisons of means
##     95% family-wise confidence level
## 
## Fit: aov(formula = log_particle_count ~ phase, data = .)
## 
## $phase
##                              diff        lwr        upr     p adj
## pouring-crushing        1.0148760  0.1042766  1.9254754 0.0255023
## transitioning-crushing -0.5455612 -1.4561606  0.3650382 0.3266477
## transitioning-pouring  -1.5604372 -2.4710366 -0.6498378 0.0003559

Indicates significant difference for pouring-crushing and transitioning-pouring

Phase Comparison for MAXCRUSH with Unit Dose Packaging

Shapiro-Wilk Test

Controlled Scenario

## # A tibble: 3 × 2
##   phase          p_value
##   <chr>            <dbl>
## 1 crushing      0.000102
## 2 pouring       0.991   
## 3 transitioning 0.0540

Indicates data does not follow a normal distribution

Shapiro-Wilk Test

Uncontrolled Scenario

## # A tibble: 3 × 2
##   phase         p_value
##   <chr>           <dbl>
## 1 crushing       0.0455
## 2 pouring        0.0567
## 3 transitioning  0.0120

Indicates data does not follow a normal distribution

Kruskal-Wallis Test

Controlled Scenario

## [1] 1.586744e-08

Indicates that at least one group is significantly different from the others

Dunns Test

Controlled Scenario

##                 Comparison         Z      P.unadj        P.adj
## 1       crushing - pouring -5.804954 6.438349e-09 1.931505e-08
## 2 crushing - transitioning -1.611984 1.069654e-01 3.208961e-01
## 3  pouring - transitioning  4.192970 2.753256e-05 8.259767e-05

Indicates significant difference for crushing-pouring and pouring-transitioning

Kruskal-Wallis Test

Uncontrolled Scenario

## [1] 9.122313e-08

Indicates that at least one group is significantly different from the others

Dunns Test

Controlled Scenario

##                 Comparison          Z      P.unadj        P.adj
## 1       crushing - pouring -5.3011625 1.150676e-07 3.452028e-07
## 2 crushing - transitioning -0.8510833 3.947231e-01 1.000000e+00
## 3  pouring - transitioning  4.4500792 8.583862e-06 2.575159e-05

Indicates significant difference for crushing-pouring and pouring-transitioning

Phase Comparison for Silent Knight

Shapiro-Wilk Test

Controlled Scenario

## # A tibble: 3 × 2
##   phase         p_value
##   <chr>           <dbl>
## 1 crushing       0.201 
## 2 pouring        0.109 
## 3 transitioning  0.0923

Indicates data follows a normal distribution

Shapiro-Wilk Test

Uncontrolled Scenario

## # A tibble: 3 × 2
##   phase         p_value
##   <chr>           <dbl>
## 1 crushing      0.00196
## 2 pouring       0.230  
## 3 transitioning 0.0596

Indicates data does not follow a normal distribution

Bartlett Test

Controlled Scenario

## 
##  Bartlett test of homogeneity of variances
## 
## data:  log_particle_count by phase
## Bartlett's K-squared = 0.030444, df = 2, p-value = 0.9849

Indicates that variance of groups are equal

Levenes Test

Controlled Scenario

## Levene's Test for Homogeneity of Variance (center = median)
##       Df F value Pr(>F)
## group  2  0.4389 0.6469
##       57

Indicates that variance of groups are equal

ANOVA Test

Controlled Scenario

##             Df Sum Sq Mean Sq F value   Pr(>F)    
## phase        2   32.1  16.050   9.026 0.000393 ***
## Residuals   57  101.3   1.778                     
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Indicates that at least one group is significantly different from the others

Tukeys HSD Test

Controlled Scenario

##   Tukey multiple comparisons of means
##     95% family-wise confidence level
## 
## Fit: aov(formula = log_particle_count ~ phase, data = .)
## 
## $phase
##                              diff         lwr        upr     p adj
## pouring-crushing        1.0986051  0.08388705  2.1133232 0.0309214
## transitioning-crushing -0.6763645 -1.69108260  0.3383535 0.2522578
## transitioning-pouring  -1.7749696 -2.78968772 -0.7602516 0.0002676

Indicates significant difference for pouring-crushing and transitioning-pouring

Kruskal Wallis Test

Uncontrolled Scenario

## [1] 6.450195e-08

Indicates that at least one group is significantly different from the others

Dunns Test

Uncontrolled Scenario

##                 Comparison         Z      P.unadj        P.adj
## 1       crushing - pouring -4.197828 2.694871e-05 8.084614e-05
## 2 crushing - transitioning  1.309683 1.903030e-01 5.709090e-01
## 3  pouring - transitioning  5.507511 3.639421e-08 1.091826e-07

Indicates significant difference for crushing-pouring and pouring-transitioning

Phase Comparison for Safe Crush

Shapiro-Wilk Test

Controlled Scenario

## # A tibble: 3 × 2
##   phase         p_value
##   <chr>           <dbl>
## 1 crushing       0.710 
## 2 pouring        0.0388
## 3 transitioning  0.0512

Indicates data does not follow a normal distribution

Shapiro-Wilk Test

Uncontrolled Scenario

## # A tibble: 3 × 2
##   phase         p_value
##   <chr>           <dbl>
## 1 crushing       0.482 
## 2 pouring        0.0675
## 3 transitioning  0.639

Indicates data follows a normal distribution

Kruskal Wallis Test

Controlled Scenario

## [1] 2.433138e-07

Indicates that at least one group is significantly different from the others

Dunns Test

Controlled Scenario

##                 Comparison         Z      P.unadj        P.adj
## 1       crushing - pouring -4.020905 5.797510e-05 1.739253e-04
## 2 crushing - transitioning  1.263325 2.064725e-01 6.194175e-01
## 3  pouring - transitioning  5.284229 1.262351e-07 3.787053e-07

Indicates significant difference for crushing-pouring and pouring-transitioning

Bartlett Test

Controlled Scenario

## 
##  Bartlett test of homogeneity of variances
## 
## data:  log_particle_count by phase
## Bartlett's K-squared = 6.2474, df = 2, p-value = 0.04399

Indicates that variance of groups are equal

Levenes Test

Controlled Scenario

## Levene's Test for Homogeneity of Variance (center = median)
##       Df F value Pr(>F)
## group  2  1.6982 0.1921
##       57

Indicates that variance of groups are unequal

Kruskal Wallis Test

Uncontrolled Scenario

## [1] 2.420254e-08

Indicates that at least one group is significantly different from the others

Dunns Test

Uncontrolled Scenario

##                 Comparison         Z      P.unadj        P.adj
## 1       crushing - pouring -4.545968 5.468328e-06 1.640499e-05
## 2 crushing - transitioning  1.014240 3.104683e-01 9.314048e-01
## 3  pouring - transitioning  5.560208 2.694531e-08 8.083593e-08

Indicates significant difference for crushing-pouring and pouring-transitioning

Tests with Untransformed Data

Boxplot Comparing Containers in Controlled and Uncontrolled Scenarios

Shapiro-Wilk Tests

Controlled Scenario

## # A tibble: 4 × 2
##   container        p_value
##   <chr>              <dbl>
## 1 cup           0.0710    
## 2 packaged      0.000563  
## 3 safecrush     0.00000126
## 4 silent knight 0.00000228

Uncontrolled Scenario

## # A tibble: 4 × 2
##   container      p_value
##   <chr>            <dbl>
## 1 cup           0.00709 
## 2 packaged      0.197   
## 3 safecrush     0.0446  
## 4 silent knight 0.000848

Kruskal Wallis Test

Controlled Scenario

## [1] 5.001664e-06

Indicates that at least one group is significantly different from the others

Dunns Test

Controlled Scenario

##                  Comparison          Z      P.unadj        P.adj
## 1            cup - packaged -4.7391656 2.146001e-06 1.287601e-05
## 2           cup - safecrush -0.4660917 6.411499e-01 1.000000e+00
## 3      packaged - safecrush  4.2730739 1.927965e-05 1.156779e-04
## 4       cup - silent knight -1.9119964 5.587665e-02 3.352599e-01
## 5  packaged - silent knight  2.8271691 4.696152e-03 2.817691e-02
## 6 safecrush - silent knight -1.4459048 1.482039e-01 8.892235e-01

Indicates significant difference for packaged-cup, safecrush-packaged, and silent knight-packaged

Kruskal Wallis Test

Uncontrolled Scenario

## [1] 8.444565e-11

Indicates that at least one group is significantly different from the others

Dunns Test

Uncontrolled Scenario

##                  Comparison         Z      P.unadj        P.adj
## 1            cup - packaged -6.459159 1.052861e-10 6.317167e-10
## 2           cup - safecrush -2.232286 2.559610e-02 1.535766e-01
## 3      packaged - safecrush  4.226874 2.369603e-05 1.421762e-04
## 4       cup - silent knight -5.008838 5.475967e-07 3.285580e-06
## 5  packaged - silent knight  1.366974 1.716334e-01 1.000000e+00
## 6 safecrush - silent knight -2.805357 5.026083e-03 3.015650e-02

Indicates significant difference for packaged-cup, silent knight-cup, safecrush-packaged, and silent knight-safecrush. Tukey showed difference between safecrush-cup

Comparing Controlled & Uncontrolled for Each Container

Shapiro-Wilk Test MAXCRUSH with Pill Cup

## # A tibble: 2 × 2
##   scenario     p_value
##   <chr>          <dbl>
## 1 controlled   0.0710 
## 2 uncontrolled 0.00709

Wilcoxon Rank Sum Test MAXCRUSH with Pill Cup

## 
##  Wilcoxon rank sum test with continuity correction
## 
## data:  particle_count by scenario
## W = 145.5, p-value = 0.1441
## alternative hypothesis: true location shift is not equal to 0

Indicates no significant difference between groups

Shapiro-Wilk Test MAXCRUSH with Unit Dose Packaging

## # A tibble: 2 × 2
##   scenario      p_value
##   <chr>           <dbl>
## 1 controlled   0.000563
## 2 uncontrolled 0.197

Wilcoxon Rank Sum Test with Unit Dose Packaging

## 
##  Wilcoxon rank sum exact test
## 
## data:  particle_count by scenario
## W = 37, p-value = 1.654e-06
## alternative hypothesis: true location shift is not equal to 0

Indicates significant difference between groups

Shapiro-Wilk Test Silent Knight

## # A tibble: 2 × 2
##   scenario        p_value
##   <chr>             <dbl>
## 1 controlled   0.00000228
## 2 uncontrolled 0.000848

Wilcoxon Rank Sum Test Silent Knight

## 
##  Wilcoxon rank sum exact test
## 
## data:  particle_count by scenario
## W = 33, p-value = 1.503e-06
## alternative hypothesis: true location shift is not equal to 0

Indicates significant difference between groups

Shapiro-Wilk Test Safe Crush

## # A tibble: 2 × 2
##   scenario        p_value
##   <chr>             <dbl>
## 1 controlled   0.00000126
## 2 uncontrolled 0.0446

Wilcoxon Rank Sum Test Safe Crush

## 
##  Wilcoxon rank sum exact test
## 
## data:  particle_count by scenario
## W = 76, p-value = 0.0005305
## alternative hypothesis: true location shift is not equal to 0

Indicates significant difference between groups

Phase Comparison for MAXCRUSH with Pill Cup

Shapiro-Wilk Test

Controlled Scenario

## # A tibble: 3 × 2
##   phase           p_value
##   <chr>             <dbl>
## 1 crushing      0.0000243
## 2 pouring       0.00550  
## 3 transitioning 0.0320

Kruskal Wallis Test

Controlled Scenario

## [1] 3.35799e-05

Indicates that at least one group is significantly different from the others

Dunns Test

Controlled Scenario

##                 Comparison          Z      P.unadj        P.adj
## 1       crushing - pouring -3.3464118 8.186472e-04 2.455942e-03
## 2 crushing - transitioning  0.9826405 3.257844e-01 9.773532e-01
## 3  pouring - transitioning  4.3290523 1.497524e-05 4.492571e-05

Indicates significant difference for crushing-pouring and pouring-transitioning

Shapiro-Wilk Test

Uncontrolled Scenario

## # A tibble: 3 × 2
##   phase           p_value
##   <chr>             <dbl>
## 1 crushing      0.00101  
## 2 pouring       0.000514 
## 3 transitioning 0.0000131

Kruskal Wallis Test

Uncontrolled Scenario

## [1] 0.001893469

Indicates that at least one group is significantly different from the others

Dunns Test

Uncontrolled Scenario

##                 Comparison         Z      P.unadj       P.adj
## 1       crushing - pouring -1.924419 0.0543020363 0.162906109
## 2 crushing - transitioning  1.611984 0.1069653693 0.320896108
## 3  pouring - transitioning  3.536404 0.0004056145 0.001216843

Indicates significant difference for transitioning-pouring. Tukey indicated significant difference for pouring-crushing

Phase Comparison for MAXCRUSH with Unit Dose Packaging

Shapiro-Wilk Test

Controlled Scenario

## # A tibble: 3 × 2
##   phase              p_value
##   <chr>                <dbl>
## 1 crushing      0.0000000216
## 2 pouring       0.000101    
## 3 transitioning 0.0000000133

Kruskal-Wallis Test

Controlled Scenario

## [1] 1.586744e-08

Dunns Test

Controlled Scenario

##                 Comparison         Z      P.unadj        P.adj
## 1       crushing - pouring -5.804954 6.438349e-09 1.931505e-08
## 2 crushing - transitioning -1.611984 1.069654e-01 3.208961e-01
## 3  pouring - transitioning  4.192970 2.753256e-05 8.259767e-05

Indicates significant difference for crushing-pouring and pouring-transitioning

Shapiro-Wilk Test

Uncontrolled Scenario

## # A tibble: 3 × 2
##   phase              p_value
##   <chr>                <dbl>
## 1 crushing      0.000000399 
## 2 pouring       0.0315      
## 3 transitioning 0.0000000836

Kruskal-Wallis Test

Uncontrolled Scenario

## [1] 9.122313e-08

Indicates that at least one group is significantly different from the others

Dunns Test

Uncontrolled Scenario

##                 Comparison         Z      P.unadj        P.adj
## 1       crushing - pouring -5.804954 6.438349e-09 1.931505e-08
## 2 crushing - transitioning -1.611984 1.069654e-01 3.208961e-01
## 3  pouring - transitioning  4.192970 2.753256e-05 8.259767e-05

Indicates significant difference for crushing-pouring and pouring-transitioning

Shapiro-Wilk Test

Controlled Scenario

## # A tibble: 3 × 2
##   phase             p_value
##   <chr>               <dbl>
## 1 crushing      0.0000368  
## 2 pouring       0.000000180
## 3 transitioning 0.000000654

Shapiro-Wilk Test

Uncontrolled Scenario

## # A tibble: 3 × 2
##   phase               p_value
##   <chr>                 <dbl>
## 1 crushing      0.00000000882
## 2 pouring       0.000599     
## 3 transitioning 0.0000000104

Kruskal Wallis Test

Controlled Scenario

## [1] 0.001129727

Indicates that at least one group is significantly different from the others

Dunns Test

Controlled Scenario

##                 Comparison         Z      P.unadj        P.adj
## 1       crushing - pouring -2.259495 0.0238526257 0.0715578771
## 2 crushing - transitioning  1.390110 0.1644954758 0.4934864274
## 3  pouring - transitioning  3.649605 0.0002626441 0.0007879323

Indicates significant difference for transitioning-pouring

Kruskal Wallis Test

Uncontrolled Scenario

## [1] 6.450195e-08

Indicates that at least one group is significantly different from the others

Dunns Test

Uncontrolled Scenario

##                 Comparison         Z      P.unadj        P.adj
## 1       crushing - pouring -4.197828 2.694871e-05 8.084614e-05
## 2 crushing - transitioning  1.309683 1.903030e-01 5.709090e-01
## 3  pouring - transitioning  5.507511 3.639421e-08 1.091826e-07

Indicates significant difference for crushing-pouring and pouring-transitioning

Phase Comparison for Safe Crush

Shapiro-Wilk Test

Controlled Scenario

## # A tibble: 3 × 2
##   phase              p_value
##   <chr>                <dbl>
## 1 crushing      0.00173     
## 2 pouring       0.000000117 
## 3 transitioning 0.0000000243

Shapiro-Wilk Test

Uncontrolled Scenario

## # A tibble: 3 × 2
##   phase         p_value
##   <chr>           <dbl>
## 1 crushing      0.00283
## 2 pouring       0.0370 
## 3 transitioning 0.587

Kruskal Wallis Test

Controlled Scenario

## [1] 2.433138e-07

Indicates that at least one group is significantly different from the others

Dunns Test

Controlled Scenario

##                 Comparison         Z      P.unadj        P.adj
## 1       crushing - pouring -4.020905 5.797510e-05 1.739253e-04
## 2 crushing - transitioning  1.263325 2.064725e-01 6.194175e-01
## 3  pouring - transitioning  5.284229 1.262351e-07 3.787053e-07

Indicates significant difference for crushing-pouring and pouring-transitioning

Kruskal Wallis Test

Uncontrolled Scenario

## [1] 2.420254e-08

Indicates that at least one group is significantly different from the others

Dunns Test

Uncontrolled Scenario

##                 Comparison         Z      P.unadj        P.adj
## 1       crushing - pouring -4.545968 5.468328e-06 1.640499e-05
## 2 crushing - transitioning  1.014240 3.104683e-01 9.314048e-01
## 3  pouring - transitioning  5.560208 2.694531e-08 8.083593e-08

Indicates significant difference for crushing-pouring and pouring-transitioning

## # A tibble: 3 × 2
##   medication p_value
##   <chr>        <dbl>
## 1 calcium     0.104 
## 2 senokot     0.0204
## 3 tylenol     0.276

## 
##  Kruskal-Wallis rank sum test
## 
## data:  particle_per_gram by medication
## Kruskal-Wallis chi-squared = 2.2924, df = 2, p-value = 0.3178

Pill Crushing

VCH Hygiene

23 May, 2025

Objective

Background

Normality Visuzlization

QQ Plot

Normality Visualization Log Transformed

QQ Plot Log Transformed

Boxplot Comparing Containers in Controlled and Uncontrolled Scenarios

Shapiro-Wilk Tests

Controlled Scenario

Uncontrolled Scenario

Bartlett Test

Controlled Scenario

Uncontrolled Scenario

Levenes Tests

Controlled Scenario

Uncontrolled Scenario

ANOVA

Controlled Scenario

Tukeys HSD Test

Controlled Scenario

ANOVA

Uncontrolled Scenario

Tukeys HSD Test

Uncontrolled Scenario

Comparing Controlled & Uncontrolled for Each Container

Shapiro-Wilk Test MAXCRUSH with Pill Cup

Bartlett Test MAXCRUSH with Pill Cup

Levenes Test MAXCRUSH with Pill Cup

T Test MAXCRUSH with Pill Cup

Shapiro-Wilk Test MAXCRUSH with Unit Dose Packaging

Bartlett Test MAXCRUSH with Unit Dose Packaging

Levenes Test MAXCRUSH with Unit Dose Packaging

T Test MAXCRUSH with Unit Dose Packaging

Shapiro-Wilk Test Silent Knight

Bartlett Test Silent Knight

Levenes Test Silent Knight

T Test Silent Knight

Shapiro-Wilk Test Safe Crush

Bartlett Test Test Safe Crush

Levenes Test Safe Crush

T Test Safe Crush

Phase Comparison for MAXCRUSH with Pill Cup

Shapiro-Wilk Test

Controlled Scenario

Shapiro-Wilk Test

Uncontrolled Scenario

Bartlett Test

Controlled Scenario

Bartlett Test

Uncontrolled Scenario

Levenes Test

Controlled Scenario

Levenes Test

Uncontrolled Scenario

Kruskal Wallis Test

Controlled Scenario

Dunns Test

Controlled Scenario

ANOVA Test

Uncontrolled Scenario

Tukeys HSD Test

Uncontrolled Scenario

Phase Comparison for MAXCRUSH with Unit Dose Packaging

Shapiro-Wilk Test

Controlled Scenario

Shapiro-Wilk Test

Uncontrolled Scenario

Kruskal-Wallis Test

Controlled Scenario

Dunns Test

Controlled Scenario

Kruskal-Wallis Test

Uncontrolled Scenario

Dunns Test

Controlled Scenario

Phase Comparison for Silent Knight

Shapiro-Wilk Test