The nursing profession is facing possibly its greatest challenge in terms of recruitment and retention. Although the number of employed nurses in New South Wales (NSW) Australia continues to gradually decline, patient turnover in our healthcare system continues to increase steadily. 1 Enrollments in undergraduate nursing courses in Australia have fallen by 11% during 6 years. 2 Furthermore, annual recruitment to the profession is projected to continue declining worldwide in the next few years. 3,4 A crucial issue in the recruitment and retention dilemma is the influence of nurses’ job satisfaction. The causal relationship between job satisfaction and staff turnover has been established and has ramifications for newly registered graduate nurses and the retention of experienced nurses. 5–8 Nurse researchers cite the incongruence among responsibility, authority and empowerment as the major occupational stressor and hence a leading cause for attrition in nursing. 6,7,9 The increasing necessity to retain qualified and experienced nurses’ highlights the need to understand and confidently measure the multiple dimensions of nurses’ job satisfaction.
Research into job satisfaction suggests that as a job attribute it is possibly the most significant, yet elusive factor in understanding worker motivation, performance, and effectiveness. 10–12 The direct and most often attributable effect of job satisfaction lies in recruitment and retention for an organization. 13–15 Indeed, often it is on these grounds alone that many researchers seek to measure job satisfaction empirically. 10,16,17 Work is an important feature for much of our adult life. Our principle choice of work can help shape our view of ourselves. 18 Several factors are commonly featured in explorations of employment. One is financial gain, which is sought from employment. 19,21 A sense of fulfillment and achievement also is viewed as a necessity. 5,19,22 Employment also relates to activity that broadens our daily life and helps to give meaning to our existence. 5,23 Although many researchers discuss job satisfaction as a single psychological construct, it is just one feature of job attitude that includes morale, personal and industrial development, and life purpose. 18,20,23–26 In the broadest sense, job satisfaction is about how we feel about our job. 27 We aim to feel good for a job well done and we endeavor to feel satisfied with all aspects of our job. If this is not possible, an employee may re-evaluate his or her position. 28
The Significance of Job Satisfaction for the Retention of Nurses
Job satisfaction is a critical factor in healthcare settings for several reasons. Strong empirical evidence supports a causal relationship among job satisfaction, staff turnover, and absenteeism. 29 The evidence suggests that when nurses’ job satisfaction is low, retention of staff also is low and staff turnover and absenteeism increases. 29 This combination of events can cause significantly lower standards in healthcare delivery. In addition, the costs of recruitment can be measured in both financial and productivity terms. Klinefelter 30 suggests that interest in nurses’ job satisfaction is often motivated solely by the organization’s cost effectiveness rather than by any concern for the individual or for the nursing profession. When new nursing staff must continually be recruited due to poor staff retention evolving from low job satisfaction, other areas (eg, budget and cost containment) become problematic. 30,31 Again, the long-term effect may be a spiralling downturn in standards of healthcare delivery due to short-term mismatches in time and care management. In addition, being short-staffed is in itself likely to lower job satisfaction and further compound the difficulties of retaining nursing staff.
This difficulty in replacing and retaining nursing staff is not a localized phenomena; rather, it is a worldwide issue as the cycle of available nursing staff reaches a new low point. 3,4 Staff shortages that evolve from low job satisfaction cannot be remedied easily when the available pool of recruits declines. Those who have left the profession due to low job satisfaction may be difficult to entice back into the profession because of their prior experience. As such, the evidence from nursing research suggests that job satisfaction continues to play a pivotal role in retaining nursing staff. 11 If nursing work is understood by potential new recruits as offering little in the way of job satisfaction, then recruitment strategies must address this critical factor.
A descriptive correlational design with a qualitative component is used in a study to explore the job satisfaction for newly graduated and experienced nurses in NSW, Australia. A longitudinal element is superimposed onto the study design. Job satisfaction is measured over two points in time (8 months apart) with the same groups of participants. The intention of nurses’ to stay in nursing was measured using a newly created Nurses’ Retention Index (NRI) at Time 2 (T2) only.
Participants were asked to complete either an electronic (via the Internet) or a paper version of a survey questionnaire. Confidentiality and anonymity was explained as well as the need for a name/address from participants for the follow-up survey. The survey took into account terminology commonly used by the nursing profession. The qualitative data from T2 administration was entered into a word processing program for initial clustering of themes.
The participants for the study were two groups of nurses differentiated by their nursing experience.
The participants for Group One (G1) were in the last semester of the Bachelor of Nursing program, at six universities in the Sydney region. The students came from many different backgrounds regarding age, country of birth, gender, and prior experience in nursing. A total of 506 of 870 student nurses (58%) approached by the researcher consented to participate in the study. The mean age of participants (n = 506) was 22 years; however, one-quarter of all participants were older than 32 years of age. Eighty-five percent of the sample (n = 427) were female and 15% of the sample were male (n = 75). This pattern of gender representation reflects the ongoing trend of a female dominated profession. Participants varied markedly on nursing experience with 58% of participating student nurses (n = 289) reporting that they had no nursing experience and 41% (n = 206) reporting a range of 1 to 20 years experience. More than half of the participants (64%) reported their country of birth as Australia (n = 289); the remainder reported being born in 50 different countries (n = 163). At Time Two (T2) of the study, G1 participants had graduated from their respective courses and had begun work in a graduate nurse program as first year registered nurses (RNs) in various hospitals in NSW. From the initial participant pool, a total of 110 graduate nurses agreed to participate in the second phase of the study.
The participants for Group Two (G2) were randomly selected from the NSW Nurses’ Registration Board database. This database is updated yearly and remains the confidential property of the NSW Nurses’ Registration Board. Two-thousand working nurses were selected from the database in 1999 to be approached by mail to participate in the study. Of these nurses, 528 (26.4%) agreed to participate. The random sample of nurses demonstrated a varying ages, experience, qualifications, and location. The age of participants ranged from 22 through to 79, with a mean age of 42.4 years, reflecting the aging nature of the Australian nursing work force. Most participants were female (94.6%). Most participants listed their country of birth as Australia (79%). The remaining 21% represented nurses from England, New Zealand, and the Philippines. The mean number of years for experience in nursing was 22 years (SD 9 years). At least 65% of participants stated their highest qualifications as a certificate, although an additional 35% of nurses stated they had attained qualifications ranging from a diploma through master’ degree. From the initial pool (n = 528) 332 (63%) experienced nurses agreed to participate at Time Two (T2).
The Index of Work Satisfaction 27 is used to measure nurses’ job satisfaction. The theoretical basis used for the Index of Work Satisfaction (IWS) is that of motivational theory as it relates to work behaviour. 27 The IWS is a two-part multidimensional job satisfaction instrument. Part B consists of a paired comparison measure utilizing 15 pairs of components based on the six dimensions of pay, Professional Status, Autonomy, Organizational Policies, Task Requirements, and Interactions. Part A consists of an attitude measure composed of the same six factors in 44 items. A seven-point Likert type scale is used in the IWS ranging from 1 (strongly agree) to 7 (strongly disagree).
Scoring the Index of Work Satisfaction
The scores from two parts of the IWS were totaled and standardized and were combined to produce an overall total weighted score, which is referred to as the index. The paired comparisons produced a rank of importance that demonstrates the six factors of importance in descending order (Table 1). The rank of satisfaction is determined from the process of totaling all scores for each factor and creating a mean score for each factor. A ranking of which factor provides more or less satisfaction for nurses is determined from these results (Table 2).
Reliabilities of the IWS Part A
There is considerable evidence of internal consistency for Part A of the IWS with more than 50 other studies reporting similar results to those reported here. The overall Coefficient alpha estimate of reliability is reported as >.90 over several studies 27, (p.93) and was found to be similar in this study. The subscale reliabilities for the IWS in this study show a higher than reported alpha score for the professional status factor (.60) than those published in the IWS manual.45 The results are greater than .70 for all other scales and are consistent with many other studies that have used the IWS. 27
Stamps reports the development of factorial validity for the IWS Part A based on exploratory factor analysis (EFA) that used a “Principal Components Factor analysis with a Varimax rotation.”27(p193) In the earlier EFA’s (before the revision to a 44 item measure) Stamps reported the finding of 12 factors accounting for 62% of the variance. 27(p.201) Two more recent EFA’s reported by users 27 demonstrate a 12-factor solution on the revised 44 item version in 1995 and 1996 that accounted for 52.4% of the variance. The results based on an EFA (PCF with Varimax rotation) for the present study identified a 10 factor solution accounting for 60% of the variance. However, the interaction factor represents two factors; nurse to nurse and nurse to doctor. By utilizing a Principle Axis Factor (PAF) technique with an Oblique rotation and by fixing items to their theoretically designated factors, the results clearly identified the hypothesized six factors. Target loadings were reasonable to substantial, ranging from .33 to .84. These results represent a substantial improvement in factor allocation and weightings from those reported by Stamps. Crossloadings occurred for the subscales of Autonomy and Interaction, whereas the factors of Professional Status, Autonomy, Organizational Policies, and Task Requirements were reported as crossloading in the studies discussed by Stamps. 27 Due to these conflicting results, the structure of the IWS was examined further by means of confirmatory factor analysis (CFA). A CFA can provide statistically stronger evidence of factor structure and can clarify specific item or factor problems.
A CFA was conducted utilizing item pairs for the six factor measure with items restricted to their proposed factors. The results of the CFA are calculated on G2 at T1, as G1 at T1 had not commenced in the workplace. The paired item analysis demonstrated factor loadings greater than 0.5 for each factor. The factor of Pay demonstrates substantial statistical evidence in both EFA and CFA. All factors are significantly correlated (>.05), although the factor of Organizational Policies and Autonomy are highly correlated at >.80. The goodness-of-fit indexes indicate an acceptable model on item paired analysis (>.90). There are no published SEM details of the IWS to compare these results. These results, combined with those of the exploratory factor analysis, support the multidimensional structure of the IWS Part A.
Qualitative Component to Survey
A qualitative component was added for the second round of the survey at the request of the participants. An extra page was provided at the end of the survey for those participants in G1 and G2 who wished to express their thoughts on any issue relating to the nursing profession. For example, participants were asked to express their thoughts to the following statement: “Please write any comments you would like to make about your nursing experience in the area below.” The narratives ranged from a sentence to several pages and were grouped according to the predominant theme of content. A content analysis was conducted to identify key issues. It was surprising to find that several nurses related their thoughts to issues raised within the job satisfaction measure despite the open structure of the qualitative component. The topic of pay was the most commonly raised theme throughout closely followed by autonomy and attrition.
Development and Testing of Nurses’ Retention Scale
Nurses’ retention was measured by a newly created eight-item index. The aim of the Nurses’ Retention Index (NRI) was to measure nurses’ intention of staying in their nursing job or leaving to find other work. Several declarative statements were generated based on the single global construct of career or job intention. The NRI contains eight declarative type items and used an eight-point forced choice Likert type scale. There is a balance of positively and negatively worded items and all negatively worded items have their scores reversed for statistical analysis.
Content and Construct Validity
The NRI was developed specifically for this study. Fourteen items were originally developed and trialed on 23 experienced nurses to assess their face and content validity. From a review of comments and scores, 10 items were selected and reviewed by three experts for their face and content validity. Eight items were chosen based the expert reviewers comments and were used at T2 for G1 and G2. The aim in developing the NRI was to create a unidimensional measure of nurses’ career intentions. Consequently, individual items within the index were examined for interitem correlation rather than at the scale level. All items were found to be statistically significantly correlated with each other. The internal consistency of the index is high across groups.
A measure of internal consistency (reliability scores utilizing Cronbach’s Alpha) was generated for each item and for the index as a whole. The results for both groups at T2 show a high level of internal consistency for the eight items and an overall index alpha of .97 for G1 and .95 for G2 (Table 3). Construct validation was pursued statistically by the use of exploratory and confirmatory factor analyses. The results show the a priori one factor was produced with one eigen value that exceeded 1.0. The first item accounted for 83% of the total variance for G1 and 72% for G2.
Confirmatory Factor Analysis and Modifications to the NRI
Confirmatory factor analysis, utilizing a one factor congeneric model, was conducted on the measurement model of the NRI. A congeneric model is defined as one in which all measures or items within the model purport to measure the same construct. 32 The results for the eight item measure (Model 1 in Table 4) show a less than acceptable model fit (>.90). Hence, a series of models were re-estimated based on the removal of items to derive an acceptable model. The basis for removal of these items was made by using a combination of modification indices, correlations, and reliability indicators that demonstrated these items to be the statistically weakest of the eight items. The removal of item 2 ‘I am considering leaving nursing to begin a different profession’ (see Model 2 in Table 4) improved the goodness of fit. However, this fit (TLI = .82) remained lower than the acceptable range of TLI > .90. The removal of item 6 ‘I want to leave the nursing profession as soon as an opportunity arises,’ in addition to the removal of item 2, again resulted in increasing the model estimation, this time to reflect an acceptable fit (see Model 3 in Table 4). This final version of the NRI comprised of six items (Model 3), demonstrated an acceptable model based on the three fit indexes for the data from group 1 and group 2 (Table 4).
Descriptive and frequency analyses of the responses to the NRI show only slight differences between groups. The results provide strong initial support for the construct validity and internal consistency of the NRI. The aim of developing the NRI was to measure the retention plans of nurses within this study. As a newly developed measure, the construct validity and internal consistency across groups was scrutinized for model acceptability. The internal consistency of the index is high across groups. The results of the coefficient alpha table provide evidence of good reliability for the index and the factor structure is confirmed by the use of confirmatory factor analysis.
The initial model consisting of eight items did not achieve an acceptable goodness of fit on at least 2 fit indices. However, re-estimation of the index by shortening the number of items used to measure nurses’ retention intentions provided an acceptable model fit. Although the NRI was only used at T2, the results of the newly constructed index have provided strong support for construct validity and internal consistency across two divergent groups of nurses. The results provide strong initial support for the psychometric properties of the NRI.
An Empirical and Interpretive Exploration of Two Job Satisfaction Components
More than 200 nurses used the opportunity to express their thoughts about the profession of nursing. The topic of pay was the most commonly raised theme, closely followed by attrition.
Pay. Although the issue of pay is not ranked as the most important in Part A of the IWS, it does rate as the least satisfying in Part B by graduates and the second least satisfying by experienced nurses at T2. Although 56% of experienced nurses claimed to be satisfied with their current pay, only 35% of graduates agreed. In addition, 88% of graduates agreed that an upgrading of pay schedules was needed in comparison to only 64% of experienced nurses at T2. These results have implications for the recruitment of new nurses.
Graduate nurses expressed their unhappiness at the level of remuneration particularly in light of their Higher Education Contribution fees (HECs), which must be repaid to the government now that they are in full-time employment.
I do really enjoy being a nurse; however, the sacrifices that we are expected to make at times is not reflected in our pay packet. Being a first-year graduate on day duty paying back HECs I am earning approximately $200 a fortnight less than as an EN on night duty. It will take me 4 years at least to make up my losses.
This concern is repeated by other graduates, for example, “Too much responsibility for too little pay. Now I have a huge HECs debt for a poor, shift work paying job.” Several experienced nurses expressed their dissatisfaction with pay by comparing their work load, level of knowledge, and responsibility to what they perceive as less qualified and professional jobs. The following quote is an example:
I do enjoy my work but when I have friends my age who are secretaries, teachers, receptionists, etc. who are on higher wages than [I am] and when I compare my responsibilities to theirs, I am very disappointed with my choice of career.
Another commonly expressed theme from graduate and experienced nurses is the perceived inequality of pay for responsibility. For example:
If top level executives make a mistake, things may go wrong and create a lot of hassle for a lot of other people, but our mistake may cause instant death. What is more important? When are we going to earn what we truly deserve?
Several participants in the experienced group expressed their dismay at what they perceive as inequitable pay for the level of qualifications and a lack of salary incentives for experience and knowledge. These sentiments are eloquently expressed by the following comment:
For many nurses, the pitiful extra in the pay packet is not worth the exacting yearly criteria it takes to retain it. Whatever qualifications I have, whatever extra education I seek, however skilled I am, the bottom line is I am paid exactly the same as someone else in my area if [he or she] has been registered the same length of time as [I]. Why can’t I be encouraged to stay in bedside nursing and become better at it?
The culmination of pay versus responsibility dissatisfaction leads to retention issues as expressed by the following comment; “How many other ‘professions’ after 25 years, pay so poorly, for such commitment knowledge and skills levels as required by nurses? Is it any wonder they’re leaving in droves!” Statistically, the issue of pay was not a significant indicator for retention in a multiple regression analysis. Comments from experienced nurses confirm this issue; “more pay is nice but not a primary problem,” and “pay has never been a big issue for me if my enjoyment of work has been high.”
Profession Status. The component of Professional Status is rated as the most satisfying by both groups. This component was highly significant (>.001) in a multiple regression analysis with retention as well as being the only significant component. The implicaitons of these findings are that the greater the satisfaction with professional status the more likely the intent to stay in nursing. There were no significant differences in between-group analyses; however, it is interesting to note that more than 90% of participants in both groups agreed with the component item regarding the importance of their job.
Both graduates and experienced nurses continue to question the notion of professionalism for nursing. Sadly, one graduate questions not just professional status but also her choice of job, saying “The terrible shift work, poor pay, under staffing and above all how degrading nursing is as a supposed “profession.” I wish I had listened to all who told me not to go into nursing!”
The attitude toward the professional status of fellow nurses is the subject of the following comment:
I feel angry and at times outraged at the way the profession is viewed by others and especially because nurses have brought it on themselves. On the one hand, we have still the ‘horizontal violence’ among what I consider ought to be colleagues, bought about by clinging to hierarchies. On the other hand,we have people twittering on about being ‘professionals’ when they don’t know the meaning of the word.
Nontraditional workplaces, such as medical clinics and community based positions for nurses leads toward positive notions of professional status as noted in the following comment; “Our opinions are valued and therefore our esteem, professionalism, and our patient care is of a high standard.”
Retention and Attrition. Many of the comments relating to retention and attrition from G1 refer to the difficulties of the transition from student to registered nurse. For example:
My first five months of my new graduate program I nearly quit nursing. Some senior nurses can be really rude to new graduates and even some senior EN’s, which makes it really hard for a new graduate. Luckily I survived within that period of time.
Also, from a mature aged graduate:
I have found that many new graduates are thrown in and expected to perform as seasoned members of staff. The lack of professional support means that many new RNs will leave the health system.
The idea that nursing can be a good career but that the transitional period is taking a toll upon new graduates is explored by the following comment from a young graduate:
I don’t enjoy going to work as much as I used to. I think if I had started work on this ward, I would not have as high opinion of nursing as I do, but I know that nursing can be better than it is on this ward, but I can understand why so many new graduates leave nursing.
The total mean score for the NRI in G2 indicates a positive intention; however, many of the written comments reflect issues of burnout, excessive stress, and a perception of a poor public image. Despite these comments, several respondents touched on the themes eloquently expressed in this final comment from an experienced nurse:
I have been in nursing now for some 20 years and for the most part have found it a rewarding and fulfilling profession. I have been privileged to meet, care for, and work with some wonderful people who have enriched my life. I was fortunate enough to train through the hospital system. I have many fond memories of that time—the friendships, sharing of life experiences, team spirit, and collegiality. Nurses care for patients, families, and each other.
The aim of measuring nurses’ job satisfaction with a multidimensional measure was to determine whether nurses are satisfied in the various components of their job. The combination of job satisfaction, retention and qualitative data has provided an empirical and personal view of how newly registered graduate and experienced nurses’ view their job satisfaction. The results indicate areas of concern in professional status, remuneration, and autonomy. Both groups of nurses claim to be most satisfied with their professional status, yet do not rate this component higher than four of the six components in importance.
It is apparent from the results that how nurses’ view themselves as professionals affects their intention to continue with their nursing career. Although the results of the retention index were similar for both groups, it was expected that new graduates retention plans would exceed experienced nurses by virtue of their enthusiasm at the beginning of their nursing career. This was not evident in the results. In fact, the graduate group rated their retention plans as less than the experienced nurse group. These results highlight the idea that the effects of dissatisfaction with various aspects of nursing work are an integral part of the puzzle of nursing shortages. These critical findings have ramifications for current strategies aimed at improving the retention of all nurses and for the recruitment of new nurses.
The author thanks Professor Paula Stamps for her generous permission to use The Index of Work Satisfaction in this study and the substantial help from Professor Maree Johnson and Associate Professor Rhonda Craven from The University of Western Sydney, Macarthur Campus.
1. Teschendorff J. The chronic cycle of exploitation. Nurs Rev. 2000; 5 (2):7.
2. Royal College of Nursing. Nursing Review.
New South Wales: APN Educational Media Pty Ltd.; 1998.
3. Borland M. Nursing numbers falling. Nurs Rev.
4. Jones D. Nursing numbers down as health crisis looms. Nurs Rev. 1998; 3 (12):1.
5. Blegen MA. Nurses’ job satisfaction: a meta-analysis of related variables. Nurs Res. 1993; 42:36–41.
6. Decker FH. Occupational and nonoccupational factors in job satisfaction and psychological distress among nurses. Res Nurs Health. 1997; 20:453–464.
7. Healy C, McKay M. Identifying sources of stress and job satisfaction in the nursing environment. Aust J Adv Nurs. 1999; 17 (2):30–35.
8. Matrunola P. Is there a relationship between job satisfaction and absenteeism? J Adv Nurs. 1996; 23:827–834.
9. Hackett RD, Bycio P. An evaluation of employee absenteeism as a coping mechanism among hospital nurses. J Occup Org Psychol. 1996; 69:327–338.
10. Cavanagh SJ. Job satisfaction of nursing staff working in hospitals. J Adv Nurs. 1992; 17:704–711.
11. Nakata JA, Saylor C. Management style and staff nurse satisfaction in a changing environment. Nurs Adm Q. 1994; 18 (3):51–57.
12. Weisman CS, Nathanson CA. Professional satisfaction and client outcomes: a comparative organizational analysis. Med Care. 1985; 23:1179–1192
13. Cumbey DA, Alexander JW. The relationship of job satisfaction with organizational variables in public health nursing. J Nurs Adm. 1998; 28 (5):39–46.
14. Kramer M, Schmalenberg C. Job satisfaction and retention. Insight for the 90’s. Nursing. 1991; 21 (3):50–55.
15. Wells GT. Influence of organizational structure on nurse manager job satisfaction. Nurs Adm Q. 1990; 14 (4):1–8.
16. Onyett S, Pillinger T, Muijen M. Job satisfaction and burnout among members of community mental health teams. J Ment Health. 1997; 6:55–66.
17. Price JL, Mueller CW. Handbook of Organizational Measurement.
Cambridge, Mass: Ballinger; 1986.
18. Herzberg F, Mausner B, Snyderman B. The Motivation to Work.
2nd ed.. New York: John Wiley & Sons; 1959.
19. Cook JD, Hepworth SJ, Wall TD, Warr PB. The Experience of Work: A Compendium and Review of 249 Measures and Their Use.
London: Academic Press; 1981.
20. Cook JD, Wall TD. New work attitude measure of trust, organizational commitment and personal need non-fulfillment. J Occup Psychol. 1980; 53:39–52.
21. Spector PE. Job satisfaction: Application, Assessment, Cause, and Consequences.
Thousand Oaks, Calif: Sage Publications; 1997.
22. Tonges MC, Rothstein H, Carter HK. Sources of satisfaction in hospital nursing practice: A guide to effective job design. J Nurs Adm. 1998; 28 (5):47–61.
23. Hackman JR, Oldham GR. Motivation through the design of work: test of a theory. Org Behav Hum Perf. 1976; 16:250–279.
24. Judge TA, Hulin CL. Job satisfaction as a reflection of disposition: a multiple source causal analysis. Org Behav Hum Decision Proc. 1993; 56:388–421.
25. Law KS, Wong C. Multidimensional constructs in structural equation analysis: an illustration using the job perception and job satisfaction constructs. J Manag. 1999; 25:143–160.
26. Locke EA. Personnel attitudes and motivation. Annu Rev Psychol. 1975; 26:457–480.
27. Stamps P. Nurses and Work Satisfaction: An Index for measurement.
2nd ed. Chicago: Health Administration Press;1997.
28. Porter LW, Steers RM. Organisational, work, and personal factors in employee turnover and absenteeism. Psychol Bull. 1973; 80:151–176.
29. Parsons LC. Delegation skills and nurse job satisfaction. Nurs Econ. 1998; 16:18–26.
30. Klinefelter G. Role efficacy and job satisfaction of hospital nurses. J Nurs Staff Dev. 1993; 9:179–183.
31. Winter-Collins A, McDaniel AM. Sense of belonging and new graduate job satisfaction. J Nurs Staff Dev. 2000; 16 (3):103–111.
32. Byrne BM. Structural Equation Modeling With LISREL, PRELIS, and SIMPLIS: Basic Concepts, Applications, and Programming.
Mahwah, NJ: L. Erlbaum Associates; 1998.
© 2002 Lippincott Williams & Wilkins, Inc.