Effective bystander cardiopulmonary resuscitation (CPR) improves the chances of survival in cases of out-of-hospital cardiac arrest.1 CPR is widely taught to community members2 3 but the quality of bystander CPR training is generally far from ideal,4 and the failure of training programmes to achieve competency3 5 and an inability to retain skills after training have been cited as the main reasons for the inadequacies of bystander CPR.6^–8

The International Liaison Committee on Resuscitation (ILCOR) recommends that CPR training should be conducted using an organised plan of implementation that targets the two ends of the age spectrum—those aged >40 years and schoolchildren.4 In addition, strategies that promote CPR training in the workplace are attractive, although >75% of out-of-hospital cardiac arrests occur at home.9 10 This type of training would therefore create safer workplaces and employees’ families would also benefit.11

In the present study we trained non-medical employees at a hospital to conduct basic life support (BLS). Subject background information, test results and survey findings were examined and factors affecting BLS skill acquisition were studied.

METHODS

Subjects

Of 1545 non-medical employees at a tertiary teaching hospital in an urban area, 1432 received BLS training. Non-medical employees were defined as all employees other than doctors, nurses and emergency medical technicians, and participants consisted of administration staff, porters, ambulance drivers and technicians (radiology, pathology, physiotherapy and others). Of these 1432 subjects, 5 refused the test and 547 employees did not complete the survey or provide informed consent concerning the use of their test scores and personal information. These 552 subjects were therefore excluded and the remaining 880 subjects were enrolled in the study.

Training

Forty-one BLS training classes were conducted from March to December 2005. The maximum number of subjects per class was 35. The lead instructor was a board certified emergency physician; the other instructors were emergency department nurses with previous BLS teaching experience. One lead instructor led all the training classes and the other instructors were routinely changed.

Each class consisted of a single 3 h session which included a lecture, practice and a testing session. The lead instructor delivered a 30 min lecture to the class using materials that included the basics and theories of BLS, and used video clips of actual CPR and slides to demonstrate BLS guidelines as issued by the American Heart Association (AHA, 2000); the 2005 AHA guideline was not available during the study period. After lectures, classes were divided into groups of five, each of which was allocated an instructor and each group then practised BLS skills on a manikin.

Assessment of skills and data acquisition

At the end of each class, all subjects were tested for BLS skill on “Little Anne” with a “Clicker” (Laerdal Medical Corporation, Korea). A “Clicker” feature signals the correct compression depth. Two experienced and trained emergency department nurses who had not participated in the training sessions assessed skills. A modified 13-item checklist was used for skill assessment and a 5-point scale was used to determine overall competency; these have been shown to be reliable (inter-rater reliability 0.87 and 0.79, respectively). A modified form of the checklist and the 5-point scale were introduced by the International Liaison Committee on Resuscitation (ILCOR).4 12 However, it should be noted that “check pulse” was omitted from the original 14-item checklist as it was not viewed to be essential for non-healthcare providers. A Korean version of the modified 13-item checklist and 5-point scale was made. The two evaluators were educated for 1 h in the use of the checklist and the 5-point scale and trained at least one other CPR class. At the end of all 41 CPR classes the two trained evaluators tested the subjects.

Before evaluation, information was obtained using a questionnaire that included the following items: age, gender, type of job, educational background and previous history of CPR training. All subjects were also asked to evaluate the level of difficulty of the training session. Subjects also agreed that the information provided could be used in future studies. The study protocol was reviewed by the Institutional Review Board of our institution.

Data analysis

Age was classified as <40 years and ⩾40 years. An age of ⩾40 years was one of two target ends of the age spectrum for BLS training recommended by ILCOR.4 Participants were classified as technicians (radiology, pathology, physiotherapist and other) or non-technicians (administration staff, porters or ambulance drivers). Educational status was classified as “college or higher” or “high school”. A previous history of CPR training was classified as “first training” or “retraining”; times lapsed since previous training were not obtained. Level of difficulty was classified as “adequate” or “difficult”.

Statistical analysis was performed using SPSS Version 13.0 (SPSS Inc, Chicago, Illinois, USA) and the Fisher exact test, Student t test and the Mann-Whitney U test were used to determine differences between groups. Univariate ANOVA and ordinal regression tests were used for multivariate analyses, and an alpha value of 0.05 was employed.

RESULTS

Subject characteristics

The demographic data of the study subjects are shown in table 1.

Overall skill acquisition

Table 2 lists the percentage of subjects who performed each skill correctly. Over 90% of the subjects correctly performed the first two practical steps of the BLS (check unresponsiveness, call for help). However, about only two-thirds performed “opening of the airway” (68.0%), and “check breathing” (67.6%) properly, and slightly more than half (54.1%) attempted at least two breaths. Most of the subjects (84.4%) administered the proper number of cardiac massages correctly, but only 64.3% of subjects used the correct hand position. The mean (SD) total score achieved was 8.66 (3.57) (maximum score 13). A minority of subjects (22.3%) performed all 13 checklist skills.

With regard to the 5-point overall competency rating, more than half of the subjects (53.6%) were rated as “not competent” or “questionably competent” and 15.6% were rated as “outstanding” (table 3).

Factors affecting skill acquisition

Based on the information gathered by the survey, the subjects were divided into groups according to the characteristics that were believed to affect the acquisition of BLS skills. These groups were then compared based on test results.

Age was found to affect the 13-point checklist and 5-point overall competency rating scores, but no statistical difference was observed between subgroups categorised by gender, job type, educational background, a previous history of CPR training or perceived level of training programme difficulty.

Age

The mean 13-checklist scores for subjects aged <40 years and ⩾40 years were 8.86 and 7.93, respectively (p = 0.002). With regard to the 5-point overall competency rating, the percentage of those aged <40 years rated as “competent”, “very good” or “outstanding” was significantly greater than those aged ⩾40 years (48.1% vs 40.1%, p = 0.038).

Multivariate analyses

Multivariate analysis was performed to eliminate the effects of confounding variables and, for the 13-point checklist and 5-point overall competency rating scores, only age was found to significantly affect skill acquisition (tables 4 and 5).

Factors affecting perceived level of training difficulty

Statistical differences in perceived level of training difficulty were found for age (p = 0.004), type of job (p = 0.021), educational background (p = 0.001) and a previous history of CPR training (p = 0.022), whereas gender failed to achieve a statistical significance (p = 0.070, table 6).

DISCUSSION

We trained a large number of non-medical hospital employees using a single 3 h BLS training programme and investigated skill acquisition. We were disappointed with the result from two points of view.

Skill acquisition was poor. Less than half of the subjects performed BLS above the “competent” level according to the 5-point overall competency test. Over 50% of subjects achieved ratings of “questionably competent” or “not competent” immediately after training. Previous studies that have used similar traditional training methods have produced similar results.3 5 13 14 Several attempts have been made to increase the effectiveness of the traditional BLS training course—for example, the simplified four-step CPR15 and the staged training method.16 These methods have been found to be at least as effective or more effective than the traditional training method in terms of skill acquisition and retention.14^–16

The other disappointing finding was that previous training experience appeared not to affect skill acquisition, whereas a previous study concluded that retraining after 6–9 months helps prevent skill deterioration.14 However, in the present study, although we did not consider time lapsed since previous training, the only merit of previous training was that subjects felt that the training was easier. It appears that once the psychomotor skills acquired during training have disappeared, previous training confers no advantage, which suggests that the reinforcement of psychomotor skills is essential while they are retained, which is perhaps why ILCOR suggest that retesting be conducted biannually.4

We conducted the 41 BLS training sessions over a 10-month period to train 1432 non-medical employees. This was a labour-intensive course that required a total of >900 man-times, which is typical of traditional training courses.17 18 On the other hand, video-based self-instruction (VSI) courses are less labour-intensive in terms of training personnel. VSI minimises the amount of didactic information not related to CPR performance,19 and some randomised controlled studies have shown that VSI is at least as effective as traditional training in terms of skill acquisition and retention.20^–22 VSI would also provide a useful means of reinforcing traditional training sessions. However, for an employee training course like ours, the mandatory certification of completion with a certain method is needed.11

Our study shows that age affects BLS skill acquisition, with those aged ⩾40 years having a poor performance in terms of the 13-point checklist and the 5-point overall competency rating. Only 40% of subjects aged ⩾40 years were rated as above “competent” according to the 5-point overall competency rating scores. Moreover, subjects aged ⩾40 years are targeted for the CPR training of laypersons,4 and we therefore suggest that a specialised BLS education course or some alternative method of training should be applied to this age group. We suggest that VSI may provide a solution because it has been reported to produce better skill performances than traditional CPR training in subjects aged ⩾40 years.23 However, skill retention rates 3 months after VSI training were reported to be only equivalent to those 3 months after traditional CPR training sessions.24

One limitation of the present study is that, although we trained 1432 subjects and evaluated skill acquisition in 1427 subjects, only 880 subjects replied to the survey so 547 subjects could not be analysed. Moreover, subjects who are more confident regarding test performance are more likely to have replied to the survey. However, when we checked the 13-point checklist scores, no mean (SD) score disparity was found between responders and non-responders (8.66 (3.57) vs 8.46 (3.55)).

The second limitation was that the retention of skills was not assessed. Skill retention is an important aspect of BLS education.4 However, skill acquisition using the traditional training method was too poor, and we should have introduced a new method of education to the BLS class immediately.

The third potential shortcoming is that the 13-item skill checklist and the 5-point scale scores were determined by instructors. Although the assessments of adequate cardiac massage depth and ventilation were determined by “Clicker” and “chest rise”, the subjective assessments of cardiac massage depth and ventilation might have been inaccurate.25

CONCLUSION

We trained a large group of non-medical employees at a hospital to perform BLS and investigated skill acquisition and identified factors that affect skill acquisition. It appears that traditional BLS training is less effective, especially in those aged ⩾40 years. Methods to increase the effectiveness of BLS training should be sought.