Accuracy of tablet splitting: Comparison study between hand splitting and tablet cutter

Abstract

Background

Tablet splitting is often used in pharmacy practice to adjust the administered doses. It is also used as a method of reducing medication costs.

Objective

To investigate the accuracy of tablet splitting by comparing hand splitting vs. a tablet cutter for a low dose drug tablet.

Methods

Salbutamol tablets (4 mg) were chosen as low dose tablets. A randomly selected equal number of tablets were split by hand and a tablet cutter, and the remaining tablets were kept whole. Weight variation and drug content were analysed for salbutamol in 0.1 N HCl using a validated spectrophotometric method. The percentages by which each whole tablet’s or half-tablet’s drug content and weight difference from sample mean values were compared with USP specification ranges for drug content. The %RSD was also calculated in order to determine whether the drugs met USP specification for %RSD. The tablets and half tablets were scanned using electron microscopy to show any visual differences arising from splitting.

Results

27.5% of samples differed from sample mean values by a percentage that fell outside of USP specification for weight, of which 15% from the tablet cutter and 25% from those split by hand fell outside the specifications. All whole tablets and half tablets met the USP specifications for drug content but the variation of content between the two halves reached 21.3% of total content in case of hand splitting, and 7.13% only for the tablet cutter. The %RSDs for drug content and weight met the USP specification for whole salbutamol tablets and the half tablets which were split by tablet cutter. The halves which were split by hand fell outside the specification for %RSD (drug content = 6.43%, weight = 8.33%). The differences were visually clear in the electron microscope scans.

Conclusion

Drug content variation in half-tablets appeared to be attributable to weight variation occurring during the splitting process. This could have serious clinical consequences for medications with a narrow therapeutic-toxic range. On the basis of our results, we recommend to avoid tablet splitting whenever possible or the use of an accurate tablet splitting device when splitting cannot be avoided.

1. Introduction

Although it is most common to use the whole tablets in therapy, they can be divided into halves (Duman et al., 2000; Verrue et al., 2011). Dividing a solid dosage form offers the advantages of ease of administration to the elderly, children or patients who have difficulty in swallowing (Duman et al., 2000), to achieve doses less than the smallest available manufactured strength and it is also being advocated as a method of reducing prescription drug costs .The cost of some medication regimens can be decreased by as much as 50% (McDevitt et al., 1998).

Uneven breaking of a tablet may result in significant fluctuations in the administered dose. This may be clinically significant for drugs with a narrow therapeutic range, such as warfarin or digoxin. For many drugs, however, especially those with long half-lives and/or a wide therapeutic range, dose fluctuations are unlikely to be clinically significant.

Unless breaking tablets where dosage is not a major issue such as vitamins or analgesics, splitting tablets is not a good idea. If breaking tablets is necessary, a special tablet splitting gadget can be used. Splitting of tablets should not be prescribed for serious medical conditions, extended-release or enteric-coated tablets and tablets without a score line.

There are many different ways to split tablets in half. One way is to purchase a tablet splitter from your local pharmacy (see Fig. 1). These tablet splitters are safe and easy to use. All you need to do is to place the tablet in the proper place and then when the splitter is closed, a steel blade cuts the tablet in halves. Some tablets are scored and have a line dividing the dose in half and may be able to be snapped in half using your fingers. Other alternatives used are splitting by hands (for scored tablets) or with scissors (for unscored tablets), or with a kitchen knife (Verrue et al., 2011).

Figure 1.

Different design of tablet splitter.

Tablets with score line allow the administration of a portion of the tablet, which can then be considered as the unit dosage of the drug. However, actual dosages of hand-split tablets may deviate by more than 20% (McDevitt et al., 1998) and it may pose a serious risk for tablet uniformity and differ in the content of the two halves resulting in high or low blood levels which may affect the cure of the disease (Duman et al., 2000; Teng et al., 2002) especially if the dose is critical in disease treatment.

Few reports compared the bioavailability and dissolution of whole vs. half of the tablets and little effort has focus on the scoring effect on the uniform tablet divisibility (Duman et al., 2000). Properly scored tablets are necessary to divide the tablets into two equal halves (Duman et al., 2000). Besides the manufacturers’ decision on tablet scoring, human factors (physical and psychological) affect the final performance of the scored tablet (Duman et al., 2000). Nonetheless, a literature review concluded that the available literature was limited to adequately address the safety of this practice (McDevitt et al., 1998).

Salbutamol tablet was used as a model in this study. Salbutamol or albuterol is a short-acting β2-adrenergic receptor agonist used for the relief of bronchospasm in conditions such as asthma (one of the most common chronic diseases in Saudi Arabia), and chronic obstructive pulmonary disease.

The aim of the current study is to investigate the drug content and weight of the split half tablet by hand vs. use of tablet cutter comparing with whole tablet for Salbutamol using the drug assay analysis.

2. Materials and methods

2.1. Materials

Salbutamol 2 mg and 4 mg tablets were studied (Table 1). This drug was chosen because it is widely used in Saudi Arabia for the treatment of asthma and its low dose (4 mg) as well as the presence of whole tablet (2 mg) which will be as standard for measurement. All other chemicals and solvents used were of pharmaceutical grade.

Table 1.

Description of salbutamol tablets studied.

Drug	Tablet description	Scored	Observationsa
Salbutamol 4 mg	Pink, non-coated, circular tablet	Yes	Minimal powdering with tablet splitter, fractured at score
Salbutamol 2 mg	Orange, non-coated, circular tablet	Yes	Minimal powdering with tablet splitter, fractured at score

^aObservation of tablet characteristics were made during the tablet splitting process.

2.2. Methods

2.2.1. Spectrophotometric scanning of salbutamol sulphate

47.8 mg of salbutamol sulphate (equal to 39.83 mg of salbutamol) was dissolved in 100 ml of 0.1 N HCl. Then the samples of resulting solution were scanned for UV absorption (in range between 200 and 400 nm) via a UV system using a double-beam spectrophotometer Shimadzu (UV-160A) and matched 1-cm optical quartz cell to determine maximum absorption wavelength.

2.2.2. Calibration curve of salbutamol sulphate in 0.1 N HCl at (225 and 276.5 nm)

A standard curve was created for salbutamol sulphate, using pure drug powder diluted to 3 known concentrations (range between 0.0096 and 0.0478 mg/ml). These standard curves were established to verify accurate analysis of the drug.

2.2.3. Scanning electron microscope

A scanning electron microscope (SEM) is a type of electron microscope that images a sample by scanning it with a high-energy beam of electrons in a raster scan pattern. The electrons interact with the atoms that make up the sample producing signals that contain information about the sample’s surface topography, composition, and other properties such as electrical conductivity. SEM studies were done by research centre of the dentistry college, KSU.

2.2.4. Weight variation

A total of 20 whole tablets were randomly selected from salbutamol 4 mg and another 20 whole tablets from salbutamol 2 mg. Ten of the 20 randomly selected tablets of salbutamol 4 mg were split in halves using a Locking Tablet Cutter (Apothecary Products, Inc.) and the other 10 tablets were split in half by hand. All 20 whole tablets from 2 mg and 40 half tablets from 4 mg were weighted using a Mettler Toledo Aj150 (Mettler Toledo, Inc., Columbus, Ohio) analytical balance. The individual weight was compared with an average weight. Not more than two of the individual weights deviated from the official standard (limit ±7.5%). Assay parameters for each drug were taken directly from USP monographs.

2.2.5. Content uniformity

First the 10 whole tablets of 2 mg and 20 half-tablets selected from 40 halves were dissolved individually using a combination of shaking and sonication techniques in 25 ml of 0.1 N HCl. Then the samples were mixed well before filtration through a membrane filter. All tablets were assayed in accordance with developed and validated spectrophotometric method for determining content uniformity for whole tablets. Assay parameters for each drug were taken directly from USP monographs. The samples of each solution were assayed for drug concentration via UV system using a spectro UV-UIS Dual beam (uvs-2800, labomed, Inc.). The drug content was quantified by calculating the concentrations from the absorbance readings obtained through UV analysis of whole and half-tablet samples.

To assess the amount and acceptability of variations in drug content and weight, several measures were calculated. The measured drug content expressed as a percent of label claim was calculated for both whole and half-tablets. Individual values for whole tablets should be in the range of 85–115% for the drugs studied (proxy USP specification for drug content). Relative standard deviation expressed as a percentage (%RSD), was calculated for whole tablets (drug content and weight) and for half-tablets (drug content and weight). The %RSD is widely used to assess the repeatability and precision of the assays used to analyse drug content. Individual medication lots for whole tablets are targeted to have a %RSD less than 6% (proxy USP specification for %RSD).

3. Results and discussion

3.1. Spectrophotometric scanning of salbutamol sulphate

The spectrophotometric scanning of salbutamol sulphate in 0.1 N HCl showed that there are two maximum absorption wavelengths at 225 and 276.5 nm (Fig. 2).

Figure 2.

Absorption spectrum of salbutamol sulphate in 0.1 N HCl.

3.2. Calibration curve of salbutamol sulphate in 0.1 N HCl at (225 and 276.5 nm)

A linear relationship between the absorbance and the concentration of salbutamol sulphate in 0.1 N HCl at (225 and 276.5 nm), in the concentration range of 0.0096–0.0478 mg/ml was observed. The regression equation is Y = 20.688 X + 0.0096 and the correlation coefficients (r) of the linear regression of the calibration curves is 0.9999.

3.3. Scanning electron microscope

The image shows that after breaking, the tablet produces some of the terrain to either increase or decrease in the fracture. It has been noted that these features exist in both cases (hand and splitter) and more clearly when using the hand. The followings are some of the SEM photographs showing this.

3.4. Weight variation test

For all whole tablets studied, measured tablet weight expressed as a percent of target weight (see Table 2) was found to fall within the proxy USP specification percentage range. The weight variation increased significantly after splitting compared to the intact tablets (see Tables 3 and 4). Measured weight expressed as a percent of target weight for half-tablets fell outside the proxy USP specification for weight for at least 3 half-tablets when cutter was used (15%), while the number increased when tablet was split by hand to 8 half-tablets (25%).

Table 2.

Weight variation test for whole tablets.

No.	Weight (g)	Difference from the mean	%RSD
1	0.1175	−0.0029	−2.39
2	0.1170	−0.0010	−0.85
3	0.1214	0.0010	0.85
4	0.1218	0.0014	1.18
5	0.1184	−0.0020	−1.64
6	0.1179	−0.0025	−2.06
7	0.1190	−0.0014	−1.14
8	0.1195	−0.0009	−0.73
9	0.1215	0.0011	0.93
10	0.1164	−0.0040	−3.30
11	0.1192	−0.0012	−0.98
12	0.1263	0.0059	4.92
13	0.1214	0.0010	0.85
14	0.1204	0.0000	0.02
15	0.1155	−0.0049	−4.05
16	0.1241	0.0037	3.09
17	0.1201	−0.0003	−0.23
18	0.1217	0.0013	1.10
19	0.1237	0.0033	2.76
20	0.1247	0.0043	3.59

Mean = 0.1204 g.

SD = 0.0029.

%RSD = 2.41.

Table 3.

Weight variation test for half tablet split by hand.

No.	Weight (g)	Difference from the Mean	%RSD
1	0.0710	0.0104	17.17
2	0.0550	−0.0056	−9.23
3	0.0640	0.0034	5.62
4	0.0588	−0.0018	−2.96
5	0.0547	−0.0059	−9.73
6	0.0631	0.0025	4.13
7	0.0623	0.0017	2.81
8	0.0600	−0.0006	−0.98
9	0.0610	0.0004	0.67
10	0.0599	−0.0007	−1.15
11	0.0571	−0.0035	−5.77
12	0.0640	0.0034	5.62
13	0.0536	−0.0070	−11.54
14	0.0606	0.0000	0.01
15	0.0550	−0.0056	−9.23
16	0.0680	0.0074	12.22
17	0.0526	−0.0080	−13.19
18	0.0672	0.0066	10.90
19	0.0593	−0.0013	−2.14
20	0.0647	0.0041	6.77

Mean = 0.0606.

SD = 0.0050.

%RSD = 8.33.

Table 4.

Weight variation test for half tablet split by splitter.

No.	Weight (g)	Difference from the Mean	%RSD
1	0.0629	0.0021	3.54
2	0.0589	−0.0019	−3.05
3	0.0598	−0.0010	−1.56
4	0.0604	−0.0004	−0.58
5	0.0625	0.0017	2.88
6	0.0588	−0.0020	−3.21
7	0.0614	0.0006	1.07
8	0.0671	0.0064	10.45
9	0.0585	−0.0023	−3.70
10	0.0625	0.0017	2.88
11	0.0574	−0.0034	−5.51
12	0.0593	−0.0015	−2.39
13	0.0633	0.0025	4.20
14	0.0590	−0.0018	−2.88
15	0.0630	0.0023	3.70
16	0.0601	−0.0007	−1.07
17	0.0550	−0.0058	−9.47
18	0.0649	0.0041	6.83
19	0.0557	−0.0051	−8.31
20	0.0645	0.0038	6.17

Mean = 0.0608.

SD = 0.0031.

%RSD = 5.11.

3.4.1. Statistical analysis

3.4.1.1. t-Test

The mean of weight variation test for half tablets split by hand is 0.0606, while the mean of weight variation test for half tablets split by splitter is 0.0608. The t-value calculated is 1.8964 with a degree of freedom of 38; meanwhile, the t-value tabulated is 2.02 at 95% confidence interval. The t-value calculated is less than the t-value tabulated so there is no different between the two means.

3.4.1.2. F-test

The SD of tablets split by hand is 0.005 and the SD of tablets split by splitter is 0.0031. Accordingly, the F-value calculated is 2.604 with a degree of freedom of 19; meanwhile, the F-value tabulated is 2.17 at 95% confidence interval. Accordingly, the F-value calculated is more than the tabulated F-value and that gives evidence of unequal population variances.

3.5. Drug content

The measured drug content was expressed as a percent of target drug content for all whole tablets and half tablets met the proxy USP specification for %RSD (see Tables 5–7). While the percentage of content variation between the two halves can reach 21.3% of total content in case of hand splitting; using the Locking Tablet Cutter gave a maximum variation of only 7.13%.

Table 5.

Salbutamol 2 mg whole tablets drug content.

No.	Weight	Content%
1	0.1175	102.01
2	0.1170	100.10
3	0.1214	103.36
4	0.1218	104.00
5	0.1184	100.89
6	0.1179	101.06
7	0.1190	103.10
8	0.1195	102.11
9	0.1215	104.94
10	0.1164	100.65
Mean	0.1190	102.22
SD	0.0020	1.59
%RSD	1.65	1.56

Table 6.

Drug content for half tablets split by hand.

No.	Weight	Content (% of label claim)
		Half tablet	Whole tablet	Range
1	0.0710	111.20	100.54	21.31
2	0.0550	89.89
3	0.0640	102.37	99.29	6.17
4	0.0588	96.21
5	0.0547	89.50	95.27	11.55
6	0.0631	101.05
7	0.0623	100.70	99.05	3.29
8	0.0600	97.41
9	0.0610	100.43	98.05	4.75
10	0.0599	95.67
Mean	0.0610	98.44
SD	0.0047	6.33
%RSD	7.69	6.43

Table 7.

Drug content for half tablets split by tablet cutter.

No.	Weight	Content (% of label claim)
		Half tablet	Whole tablet	Range
1	0.0629	94.55	95.85	2.60
2	0.0589	97.15
3	0.0598	96.90	98.25	2.70
4	0.0604	99.59
5	0.0625	99.30	98.06	2.50
6	0.0588	96.81
7	0.0614	97.02	100.58	7.13
8	0.0671	104.15
9	0.0585	94.36	97.26	5.80
10	0.0625	100.16
Mean	0.0613	98.00
SD	0.0026	2.91
%RSD	4.28	2.97

3.5.1. t-Test

The mean of drug content for half tablets split by hand is 98.44, while the mean of drug content for half tablets split by cutter is 98.00. The t-value calculated is 0.201485 with a degree of freedom of 18. The t-value tabulated is 2.1 at 95% confidence interval. The t-value calculated is less than the t-value tabulated so there is no different between the two means.

3.5.2. F-test

The SD for the tablets split by hand is 6.33, and the SD for tablets split by splitter is 2.91. Accordingly, the calculated F-value is 4.7317 with a degree of freedom of 9; meanwhile, the tabulated F-value is 3.18 at 95% confidence interval. Accordingly, the F-value calculated is more than the tabulated F-value and that gives evidence of unequal population variances.

4. Conclusion

Tablet splitting may not have adverse clinical consequences and can reduce costs for both patients and institutions (Verrue et al., 2011), but using a whole tablet is the safest way to ensure accurate dosing. However, not all formulations are suitable for splitting, and even when they are, it may lead to dose deviations. This could have serious clinical consequences for medications with a narrow therapeutic-toxic range. On the basis of our results, which demonstrated that using accurate tablet cutter is superior to hand splitting, we recommend the use of an accurate splitting device when splitting cannot be avoided (i.e. for example when the prescribed dose is not commercially available, or when there is no alternative formulation, such as a liquid). Nursing home staff performing the splitting should also be educated in splitting as accurately as possible, and should be aware of the possible clinical consequences of dose deviations. As for policy implications, we concur with previous scientific recommendations (Teng et al., 2002) that manufacturers make it possible to avoid splitting, by introducing a wider range of tablet doses or liquid formulations.