The Usefulness of Nursing Languages to Communicate a Clinical Event

Loading....

Loading....

Loading....

Loading....

Loading....

Full text

(1)

C O N T I N U I N G

E D U C A T I O N

The Usefulness of

Nursing Languages to

Communicate a

Clinical Event

JANE M. CARRINGTON, PhD, RN

The electronic health record (EHR) has been recom-mended by a variety of sources, for example, the Institute of Medicine, as a means to increase patient safety.1 Stan-dardized nursing languages have also been suggested; however, their usefulness has not been assessed. The EHR with embedded standardized nursing languages (EHRSNL) potentially strengthens nursing documenta-tion from paper-based systems in two very important ways. First, the EHR has been associated with more ac-curate and timely documentation. This documentation generally includes evaluating the effectiveness of care, describing patients’ responses to interventions, commu-nicating patient status, and meeting legal documentation requirements.2–4Second, standardized nursing languages were created to enhance nursing documentation by in-creasing clarity and reducing ambiguity.5

Nurses are the primary users of the EHRSNL as they enter information each shift and when a clinical event or change in a patient’s condition occurs. There are lim-ited examples of research exploring nurse-to-nurse com-munication using standardized nursing languages. This article reports on research seeking to increase our un-derstanding of nurses’ perceptions of the strengths and limitations of standardized nursing languages as part of the EHR to communicate patient status associated with a clinical event.

BACKGROUND

Standardized nursing languages are used by nurses to describe their care.6Examples of approved nursing lan-guages include NANDA, NIC, and NOC.7,8 Research exploring the usefulness of standardized nursing lan-guages has suggested that there are four ways lanlan-guages can improve healthcare: (1) support nursing data

CIN: Computers, Informatics, Nursing&Vol. 30, No. 2, 82–88&CopyrightB2012 Wolters Kluwer Health | Lippincott Williams & Wilkins

The purpose of this study was to explore nurses’ perceptions of the strengths and limitations of standardized nursing languages in the elec-tronic health record to communicate a clinical event. Limited examples of research exist explor-ing the effectiveness of the electronic health record with embedded standardized nursing languages as a communication system. There-fore, their effect of standardized nursing lan-guages on nurse-to-nurse communication remains largely unknown. Data from a larger study were analyzed using qualitative content analysis. Fifty-seven thematic units represented nurses’ perceptions of the electronic health record with NANDA, NIC, and NOC for docu-menting and retrieving patient information asso-ciated with a clinical event. These thematic units were further analyzed, and three categories emerged: language comprehension, inexact-ness of the languages, and language useful-ness. Standardized nursing languages were perceived to support planning care but also posed semantic challenges and fostered inac-curacies in patient information. Standardized nursing languages may constrain nurse-to-nurse communication related to a clinical event. For languages to support nurse-to-nurse communi-cation and avoid potential safety issues, facilities must deal with inaccuracies and semantic mis-understandings to provide safe patient care.

K E Y W O R D S

Electronic health record&Failure to rescue& Nurse-to-nurse communication& Standardized nursing languages

Author Affiliation: College of Nursing, University of Colorado Anschutz Medical Campus.

This research was supported by the Department of Veterans’ Affairs. Portions of the data in this article were presented at the Western Institute of Nursing Research, April 2009.

The authors have disclosed that they have no significant rela-tionship with, or financial interest in, any commercial companies per-taining to this article.

Corresponding author: Jane M. Carrington, PhD, RN, Mail Stop C288-19,13120 E 19th Ave, Room 4227, Aurora, CO 80047 (jane. carrington@ucdenver.edu).

(2)

comparisons for benchmarking at local, regional, national, and international levels; (2) provide valuable information to the healthcare organization regarding patient care for administrative decision making; (3) de-termine patient acuity by extracting data from a clinical database; and (4) improve communication among health-care team members through the use of consistent terms for the description of assessments, interventions, and outcomes.5,9–13

Research applicable to both paper-based and electronic documentation systems has suggested that standardized nursing languages have limitations. Standardized nursing languages have been reported to be difficult to use be-cause of the lack of complete alignment between terms nurses traditionally use in documentation and the terms used in standardized nursing language.9,11,14,15 Further-more, it has been suggested that standardized nursing languages reduce the ‘‘individualized’’ focus of patient documentation.16Standardized nursing languages there-fore may not fully capture the subtle changes in patient status needed by nurses to accurately describe patient care and patient outcomes.

This study addresses standardized nursing languages as a vehicle for communication of a clinical event be-tween members of the healthcare team, by examining nurse-to-nurse communication in terms of the perceived strengths and limitations of electronic documentation supported by standardized nursing language.

Significance of Nurse-to-Nurse

Communication

Effective communication between nurses during patient handoffs has the potential to decrease risks to patient safety. Failure to rescue, a risk to patient safety, is de-fined as patient deaths associated with a complication from treatment.17Antecedents to failure to rescue are high patient-to-nurse staffing ratios, chaotic work en-vironment, nurse work dissatisfaction, and ineffective nurse-physician collaboration.18–20Nurse-to-nurse com-munication as an antecedent to failure to rescue is not well understood. This study explored nurse-to-nurse com-munication by eliciting nurses’ perceptions of standard-ized nursing languages as a communication system, when entering or retrieving patient information in an EHR as-sociated with a clinical event. A clinical event is defined as an unexpected change in patient condition that does not result in a patient transfer and is not associated with a nursing protocol.21

Conceptual Framework

Elements of Information Theory served as the conceptual framework for this study.22These elements are informa-tion source,device,destination,redundancy,probability,

and noise. Theinformation sourceproduces the message using adevice consisting of a transmitter, channel, and receiver. The message is intended for thedestinationor receiver.

Redundant or repetitious messages are often sent using both verbal and written communication. Redundancy may prevent errors through duplication of the message.23,24 Nurse-to-nurse communication, nurse documentation, and change of shift report are inherently redundant. Con-tent from patient assessments and progress notes are of-ten repeated verbally during nurse change of shift report. Redundancy increases information.Probability, a mea-sure of predictability, allows the receiver to act on the information received without presuming more than what is transmitted.25

Noise, on the other hand, may have occurred when nursing documentation and change of shift report do not contain useful or understandable information. Noise dis-rupts communication and is evident when the message either does not reach the intended destination (message may be lost) or reaches the destination and is not under-standable.26Noise decreases information.

For this research, the information source is the document-ing nurse, and the device is the nursdocument-ing languages embedded in the EHR. The message destination is the receiving nurse. Redundancy and noise were operationalized for this research as the nurses’ perceptions of the strengths and limitations of the nursing languages, respectively.

METHODS

Design

This research is part of a study on nursing documenta-tion and has been previously described in detail.21,27In this study, documenting and receiving nurses were in-terviewed using a semistructured interview format. Thirty-seven nurses (20 documenting and 17 receiving nurses) from two sites were interviewed for the larger study. This report involves only the site using standard-ized nursing languages (EHRSNL), NANDA, NIC, and NOC. Text from 18 interviews (10 documenting and eight receiving nurses) was reviewed.

Qualitative content analysis was used to analyze the interview text. Nurses were interviewed from December 2007 to January 2008. Institutional review boards from the University of Arizona and the research sites granted permission to perform the study.

Setting

Nursing diagnosis (NANDA), nursing interventions (NIC), and nursing outcomes (NOC) were embedded in the EHR. The researchers approached the EHR at the con-ceptual level; therefore, the system name is intentionally

(3)

not identified. Inclusion criteria for informants were RNs who (1) worked full time on the medical, surgical, or telemetry nursing unit; (2) had experience using the electronic nursing documentation system for at least 3 months; (3) understood and spoke English; and (4) cared for or assumed care for a patient who experienced a clini-cal event within the past 24 hours.

Procedures

The researcher learned of a clinical event through com-munication with the charge nurses or other staff nurses on daily visits to the nursing units. When informed of a clin-ical event, the researcher approached the documenting nurse (RN caring for the patient who experienced the clinical event) and receiving nurse (RN receiving infor-mation to continue care) to participate in the study. In-terviews were done within 8 to 12 hours after the clinical event. Documenting and receiving nurses were recruited as individuals, rather than dyads.

Interviews took place in an isolated area selected by the participant and required 20 to 30 minutes to complete. Each interview was digitally recorded, transcribed verba-tim, and reviewed for accuracy prior to analysis. Data col-lection was discontinued when saturation was reached or no new themes emerged. Documenting nurses were asked how the nursing languages made it easy and/or difficult to document the clinical event and receiving nurses how easy and/or difficult to learn of the clinical event.

Sample

Eighteen nurses (10 documenting and eight receiving nurses) met the inclusion criteria and consented to par-ticipate in the study. Participants included six men and 12 women. Nurses had worked in a hospital for a mean of 10.33T9.5 years (range, 1–30 years). The participants had a mean of 3.04T 2.37 years (range, 0.75–9.50 years) of experience working on the nursing unit. Nurses used the EHR for 5.94T4.45 years (range, 1.0–16.0 years). Nurses averaged more than 5 years (5.15 T 5.45 years) of ex-perience using nursing languages (range, 0.75–20 years).

Eleven clinical events were captured for this study. The most frequent were changes in the level of mental status and drop in hemoglobin and hematocrit (two events each). Seven additional clinical events occurred once: fever, pain, fall, seizure activity, patient was emotional, low blood pressure, and intravenous catheter too small.

Analysis

Qualitative content analysis was used to view data from its smallest unit or data bit, to the thematic unit or a data-organizing element. For this research, 99 pages

of text data were analyzed. Fifty-seven thematic units specific to nursing languages emerged. The thematic units were then organized into categories and subcategories, and frequencies of thematic units and nursing citing themes were calculated.28,29 Categories and subcategories were reviewed by two experts, doctorally prepared informatics and nursing systems researchers with qualitative research experience, until 100% agreement was achieved.

Calculating the frequency of the number of thematic units (t) and number of nursing citing (n) the theme can provide some insight into the nurses’ perceptions. Re-dundancy of data, or when themes are repeated by the same participant or many participants beyond satura-tion,30made it difficult to determine the importance of the thematic units from the simple frequencies. There-fore, ‘‘degree of redundancy’’ (DOR) statistic was cal-culated based on the work of Miles and Huberman.31,32 The DOR, as previously described, was used to discuss proportional relationships among the number of emer-gent thematic units and the proportion of nurses who cited the theme.27The DOR was calculated by (1) iden-tifying the frequencies of the thematic unit (t) and the nurse citing each theme (n); (2) calculating the theme/ sample ratio or (t/n), by dividing the frequency of the-matic units (t) by the number of nurses citing (n) the thematic unit; and (3) proportioning the sample by di-viding the number of nurses citing (n) the thematic unit by the total sample size (number of nurses interviewed) (N) or (n/N). The final values were multiplied to arrive at the DOR calculation (t/nn/N). A DOR of 0.50 or greater was used to represent a strong category or sub-category. Three categories emerged from 57 thematic units: language comprehensiveness, inexactness of the languages, and language usefulness.

RESULTS

Thematic units were organized by similarities and then categorized. Both categories, language comprehensiveness and inexactness of the languages, were large and further organized into three subcategories: professional separa-tion, care planning, and ease of use (language comprehen-siveness) and lacks descriptiveness, fosters inaccuracies, and semantics (inexactness of the languages). The category language usefulness did not contain subcategories. Table 1 shows the thematic units that led to the organization of categories and subcategories.

Language comprehensiveness accounted for 23 the-matic units or 40% of the total 57 thethe-matic units. The strongest subcategory was care planning (DOR of 0.80) (Table 2). The category inexactness of the languages was the largest category, accounting for 29 thematic units or 50% of the total, and had two strong subcategories: fos-tering inaccuracies (DOR of 0.66) and semantics (DOR

(4)

of 0.57). The category language usefulness contained thematic units describing potential solutions for improving functionality of the nursing languages. This was a weak category accounting for less than 1% of the thematic units (five of 57).

Comparison of Documenting and

Receiving Nurse Data

Within language comprehensiveness (Table 3), neither documenting nor receiving nurses perceived

profes-sional separation or ease of use as important (DOR e0.50). Documenting nurses, however, perceived care planning as more important (DOR 0.99) than the receiving nurses (DOR 0.49). Within the category inexactness of the languages, documenting and receiv-ing nurses differed. While neither documentreceiv-ing nor receiving nurses perceived lacks descriptiveness as important, documenting nurses perceived semantics (DOR 1.00) and receiving nurses perceived fosters inaccuracies (DOR 1.00) as issues with nursing lan-guages. Semantics was a not identified as an issue for receiving nurses. Only documenting nurses identified T a b l e 1

Categories, Subcategories, and Thematic Unit Exemplars by Documenting and Receiving Nurses

Category Subcategory Thematic Unit Exemplar Attribute

Language

comprehensiveness

Professional separation For billing and separating ourselves

from medicine

Documenting nurse

We are all talking the same problems Receiving nurse

Planning care The care plan is complete and individualized Documenting nurse

It’s easy to compare interventions and evaluations with patient status

Receiving nurse

Ease of use I can document using the language Documenting nurse

The language isn’t difficult Documenting nurse

Inexactness of the languages

Lacks descriptiveness Not specific to nursing Documenting nurse

You want to put decreased cardiac output, but you know it won’t be that specific

Receiving nurse

Fosters inaccuracies Some have their own interventions Receiving nurse

Without the list, people make up their own problems

Receiving nurse

Semantics Nurses ask other nurses, ‘‘What is this

supposed to mean?’’

Documenting nurse

This is our language, but who talks that way? Documenting nurse

Language usefulness (No subcategory) Put it in common language Documenting nurse

We need a reference to look at to see some terms for the same incident

Documenting nurse

T a b l e 2

Categories, Subcategories, and Degree of Representation in 18 Nurses

Category Subcategory andt

No. of Nurses Citing (n) Frequency of Theme/Nurses Citing (t/n) Proportion of Nurses Citing and

Sample (n/N) DOR (Degree of Redundancy) (t/n)(n/N) Language comprehensiveness Professional separation (t= 4) 2 2.00 0.11 0.22 Care planning (t= 14) 12 1.20 0.67 0.80 Ease of use (t= 5) 4 1.25 0.22 0.28 Inexactness of the languages Lacks descriptiveness (t= 7) 5 1.40 0.28 0.39 Fosters inaccuracies (t= 12) 8 1.50 0.44 0.66 Semantics (t= 10) 3 3.33 0.17 0.57 Language usefulness t= 5 3 1.67 0.17 0.28 Total (t) 57

(5)

potential solutions to improve nursing languages within the category language usefulness.

DISCUSSION

Based on this research, there are strengths and limita-tions to using standardized nursing languages. Partici-pants generally perceived standardized nursing languages support planning care but pose semantic challenges and foster inaccuracies in patient information. When applied to Information Theory, the value of standardized nursing languages facilitating communication (redundancy) was planning care. The task of planning care is assisted by having a standard language to organize care. Standard-ized nursing languages, however, restricted communica-tion (noise) through fostering inaccuracies and semantic challenges. This type of noise is not only a potential prob-lem for communication of a clinical event, but is also di-rectly related to patient safety. For example, if a nurse is unable to clearly receive communication related to an event, appropriate intervention may be affected, thereby delaying effective care.

When applied to the conceptual framework, docu-menting nurses perceived the standardized nursing lan-guages (device) supporting communication through care planning while constraining communication through chal-lenges with semantics. Receiving nurses did not perceive the standardized nursing languages as supporting commu-nication; rather, the languages confounded communica-tion through fostering inaccuracies of patient informacommunica-tion. No receiving nurse identified the language as useful (DOR 0.0). The differences in perceptions between the docu-menting and receiving nurses may be accounted for when considering their roles with documentation and commu-nication. Documenting nurses are primarily responsible for planning care to facilitate continuity of care. Receiving nurses, on the other hand, are responsible for taking and

processing that information. If receiving nurses see inac-curacies in the documentation, this poses substantial risk to patient safety through misinterpretation of the signifi-cance of the information for timely intervention.

Research has suggested that standardized nursing lan-guages do not capture patient care in an individualized manner.16However, this was not supported by the find-ings since planning care emerged as a strong subcategory by documenting nurses. This may suggest that standard-ized nursing languages facilitated individualstandard-ized care planning associated with a clinical event.

Consistent with the literature, both documenting and receiving nurses perceived the standardized nursing lan-guages as not easy to use.9,11,14,15Documenting nurses stated that the languages were not ‘‘regular English’’ and were based on the ‘‘legal language.’’ One could not ac-count for their dissatisfaction with the standardized nurs-ing languages because of inexperience. For the most part, nurses in this study had considerable nursing experience as well as experience with the EHR.

From the receiving nurse’s perspective, standardized nursing languages fostered inaccuracies in patient infor-mation. They stated the languages were difficult to ‘‘fit’’ to the clinical event and stated that nurses created work-arounds by ‘‘making up’’ their own problems or inter-ventions. This practice was done, as one nurse stated, ‘‘to fit a round peg into a square hole.’’ This need to circumvent the standardized nursing languages is of concern and suggests the need to systematically reeval-uate the benefit of the standardized nursing languages and its use by nurses.

Study Limitations

Despite the adherence to elements of rigor, this study was limited in that the nurses were not asked how they used the nursing languages to communicate a clinical event. Nurses were asked to identify the strengths, limitations, T a b l e 3

Comparison of Documenting (n = 10) and Receiving Nurses (n = 8)

Documenting Nurses Receiving Nurses

t/n n/N DOR t/n n/N DOR Comprehensiveness Professional separation 3.00 0.10 0.30 1.00 0.12 0.12 Care planning 1.11 0.90 1.00 1.33 0.37 0.49 Ease of use 1.50 0.20 0.30 1.00 0.25 0.25 Inexactness Lacks descriptiveness 1.33 0.30 0.40 1.50 0.25 0.37 Fosters inaccuracies 1.00 0.40 0.40 2.00 0.50 1.00 Semantics 3.33 0.30 1.00 0.00 0.00 0.00 Usefulness 1.67 0.30 0.50 0.00 0.00 0.00

(6)

and potential solutions to improving the nursing lan-guages. Furthermore, nurses were not asked to what degree the EHR or nursing languages were used during nurse-to-nurse communication of a clinical event.

CONCLUSION

The implications of these findings are that standardized nursing languages may constrain nurse-to-nurse commu-nication of a clinical event. Nurses reported the use of workarounds to cope with the barriers using the stand-ardized nursing languages. The workarounds included using their own interventions or diagnoses. The use of workarounds calls into question the usefulness of databases constructed to facilitate decision making and benchmarking using standardized nursing languages. Therefore, for the EHRSNL to be entirely useful, these inadequacies must be dealt with, and the workarounds need to be evaluated.

Clinical events, as used and defined for this study, were deemed to be possible indicators of potential complica-tions. Clinical events captured for this study align well with the work of Needleman and colleagues,33who de-termined the common complications leading to failure to rescue events: pneumonia, shock or cardiac arrest, upper gastrointestinal tract bleeding, sepsis, or deep vein throm-bosis. For example, a drop in hemoglobin and hematocrit and low blood pressure align with bleeding, while fever and pain align with pneumonia, deep vein thrombosis, or sepsis. This suggests that ineffective nurse-to-nurse com-munication of clinical events may be viewed as a precursor to failure to rescue incidents and patient complications.

If the communication device (nursing languages) is not understood by the sender and recipient, then communi-cation is ineffective. Based on this study, albeit with a limited sample, we have some indication that while stan-dardized nursing languages can facilitate planning care by the sender, they also interfere with communication be-cause of inaccuracies of patient information and lack of semantic understanding. The data suggest that nurses perceived the inaccuracies in the patient information related to the lack of fit of the languages.

How the nursing languages affect the receiving nurses’ ability to continue care remains unknown. Further research is needed to explore the relationship between nurse-to-nurse communication and failure-to-rescue events and how stan-dardized nursing languages are used in such situations. The impact of nursing languages on nurse-to-nurse communica-tion and the subsequent relacommunica-tionship to patient safety pro-vide a fertile field for nursing research.

Acknowledgments

The author thanks Joyce Verran, PhD, RN, FAAN, and Suzanne Lareau, MS, RN, FAAN, for their thoughtful

review and constructive feedback in preparation of this manuscript.

REFERENCES

1. Institute of Medicine.Crossing the Quality Chasm. Washington, DC: National Academies Press; 2001.

2. Edelstein J. A study of nursing documentation. Nurs Manage. 1990;21(11):40–46.

3. Nahm R, Poston I. Measurement of the effects of an integrated, point-of-care computer system on quality of nursing documenta-tion and patient satisfacdocumenta-tion.Comput Nurs. 2000;18(5):220–229. 4. Newfield JS. Documentation: focusing on better rather than more.

Home Healthc Manage Pract. 2006;18(3):247–249.

5. Lunney M. (2006). Helping nurses use NANDA, NOC, and NIC.

J Nurs Adm. 2006;36(3):118–125.

6. Keneen G. Use of standardized nursing language will make nursing visible.Mich Nurs. 1999;72(2):12–13.

7. NANDA International Defining the Knowledge of Nursing. http://www.nanda.org/. Accessed December 2, 2010.

8. The Center for Nursing Classifications. http://www.ncvhs.hhs.gov/ 970416w6.htm. Accessed December 2, 2010.

9. Henry S, Warren J, Lange L, Button P. A review of major nursing vocabularies and the extent to which they have the characteristics required for implementation in computer-based systems. J Am Med Inform Assoc. 1998;5:321–328.

10. Lunney M, Delaney C, Duffy M, Morrhead S, Welton J. Advocat-ing for standardized nursAdvocat-ing languages in electronic health records.

J Nurs Adm. 2005;35(1):1–3.

11. Kim H, Harris M, Savova G, Chute C. Content coverage of SNOMED-CT toward the ICU nursing flowsheets and the acuity indicators.Stud Health Technol Inform. 2006;122:722–726. 12. Clancy TR, Delaney C, Morrison B, Gunn J. The benefits of standardized nursing languages in complex adaptive systems such as hospitals.J Nurs Adm. 2006;36(9):426–434.

13. Dennis KE, Sweeney PM, Macdonald LP, Morse NA. Point of care technology: impact on people and paperwork.Nurs Econ. 1993;11(4):229–248.

14. Campbell J, Carpenter P, Sneiderman C, Cohn S, Chute C, Warren J. Phase II evaluation of clinical coding schemes: completeness, taxonomy, mapping, definitions, and clarity. J Am Med Inform Assoc. 1997;4:238–251.

15. Harris M, Graves J, Solbrig H, Elkin P, Chute C. Embedded structures and representation of nursing knowledge.J Am Med Inform Assoc. 2000;7:539–549.

16. Lee T, Yeh CH, Ho LH. Application of a computerized nursing care plan system in one hospital: experiences in ICU nurses in Taiwan.J Adv Nurs. 2002;39(1):61–67.

17. Silber JH, Williams SV, Krakauer H, Schwartz JS. Hospital and patient characteristics associated with death after surgery: a study of adverse occurrence and failure to rescue.Med Care. 1992;30(7): 615–629.

18. Aiken LH, Clarke SP, Sloane DM, Sochalski JA, Silber JH. Hospital nurse staffing and patient mortality, nurse burn-out, and job dis-satisfaction.J Am Med Assoc. 2002;288(16):1987–1993. 19. Boyle SM. Nursing unit characteristics and patient outcomes.Nurs

Econ. 2004;22(3):111–119.

20. Clark SP, Aiken LH. Failure to rescue.Am J Nurs. 2003;103(1): 42–47.

21. Carrington JM. The effectiveness of the electronic health record with standardized nursing languages for communicating patient status related to a clinical event, University of Arizona, 2008 [doctoral dissertation].Diss Abstr Int. 2008;69(03).

22. Shannon CE. The mathematical theory of communication. In: Shannon CE, Weaver W, eds.The Mathematical Theory of Commu-nication. Chicago, IL: University of Illinois Press; 1967:31–125. 23. Campbell J.Grammatical Man: Information, Entropy, Language,

and Life. New York, NY: Simon & Schuster, Inc; 1982. 24. Pierce JR.An Introduction to Information Theory Symbols, Signals,

(7)

25. Clover TM, Thomas JA.Elements of Information Theory. 2nd ed. Hoboken, NJ: John Wiley & Sons, Inc, Publication; 2006. 26. Reza FM.An Introduction to Information Theory. New York, NY:

Dover Publications; 1994.

27. Carrington JM, Effken JA. Strengths and limitations of the electronic health record for documenting clinical events.Comput Inform Nurs. 2011;29(6):360–367.

28. Downe-Wanboldt B. Content analysis: method, applications, and issues.Health Care Women Int. 1992;13:313–321.

29. Krippendorff K.Content Analysis: An Introduction to Its Method-ology. Thousand Oaks, CA: Sage Publications; 2004.

30. Lincoln Y, Guba E.Naturalistic Inquiry. Newbury Park, CA: Sage Publications; 1985.

31. Miles MB, Huberman AM.Qualitative Data Analysis A Source-book of New Methods. Beverly Hills, CA: Sage Publications; 1984.

32. Miles MB, Humberman AM. An Expanded Sourcebook Qual-itative Data Analysis. 2nd ed. Thousand Oaks, CA: Sage Publications; 1994.

33. Needleman J, Buerhaus P, Mattke S, Stewart M, Zelevinsky K. Nurse-staffing levels and the quality of care in hospitals.N Engl J Med. 2002;346(22):1715–1722.

Figure

Updating...

References

Updating...

Related subjects :