ONS Patient Outcomes Survey 2004
ONS Patient Outcomes Survey
Conducted October 2004
Gail Mallory, PhD, RN, CNAA
Robi Thomas, MS, RN, AOCN, CHPN
Nurses are challenged to articulate and document the quality of their contributions to the health outcomes of the patient. ONS has as its mission, the promotion of excellence in oncology nursing and quality cancer care. An ONS strategic goal is to drive quality care through education, research, leadership, and advocacy. The ONS Strategic Plan for 2003-2005 speaks to assuming leadership for quality oncology care in the United States. Specifically, ONS mentions generating public awareness of the specific contributions that nursing and nursing research make to patient outcomes. The Strategic Plan additionally indicates the need to generate and disseminate new knowledge to identify key oncology nursing sensitive patient outcomes for practice, research, educational programming, and policy. The results from this survey of oncology nurses will complement the hospital level information on cancer patient outcomes from Sochalski's study and provide direction for the Oncology Nursing Society's education and research efforts regarding patient outcomes.
The ONS Patient Outcomes Survey was designed as a survey research project to gather basic descriptive information regarding oncology nurses' perceptions of oncology nursing-sensitive patient outcomes. The ONS Patient Outcomes Survey was designed to gather information regarding the impact, value, use, and education needs of oncology nurses related to oncology nursing-sensitive patient outcomes. Oncology nursing-sensitive patient outcomes are defined as "the consequences or effects of nursing interventions that result in changes in patients' health status, behavior, and problem resolution. The focus is on symptoms, functional status, safety/preventable adverse events, psychological distress, and economic outcomes." Secondary aims of the survey included to: (a) determine similarities and differences in responses to the ONS Patient Outcomes Survey of ONS members from Pennsylvania and a national sample of ONS members and (b) determine similarities and differences in responses to the ONS Patient Outcomes Survey of ONS members with e-mail (Internet) access and ONS members without e-mail (Internet) access (do ONS members with e-mail access respond differently than ONS members without e-mail access).
In October 2004, there were 30,629 ONS members who met the selected membership categories of employed full-or part-time; patient, administration or education functional area; primary position of staff nurse, manager, clinical nurse specialist, nurse practitioner, educator, clinical trials nurse, or case manager; and worked in medical oncology, radiation oncology, surgical oncology or bone marrow transplant. Of these, 24,263 had e-mail addresses in the membership database. There were a total of 1,590 ONS members with PA addresses (1,178 with e-mail addresses and 412 without e-mail). A random sample of ONS members (7,500) with e-mail was selected. A random sample of 1,000 ONS members without e-mail was selected with the goal of a response rate of 30% or 300 responses, which is the modal response rate to complex ONS U.S. Mail surveys. All of the PA members with e-mail and 100 randomly selected PA members without e-mail were included in the sample prior to the random selection of the national ONS members. The PA sample size was based on the goal of achieving a large enough sample from the PA ONS members to conduct statistical analyses to answer the secondary aim of the study. Two e-mail messages were sent to the sample with e-mail addresses (n = 8,678). The US Mail distribution occurred to the sample without e-mail addresses (n = 1,100) in early October 2004 and a postcard reminder was sent two weeks later.
The survey instrument/questionnaire "Oncology Nursing Society (ONS) Patient Outcomes Survey" was developed utilizing the oncology nursing sensitive patient outcomes and issues related to outcomes utilization identified in the literature and by the 2003 Oncology Nursing Sensitive Patient Outcomes Project Team. The questionnaire is designed to provide descriptive information from the subject (their perceptions) regarding the impact, value, use, and education needs related to oncology nursing patient outcomes, along with demographic information about the subject's professional role. The instrument was pilot tested at the ONS Mentorship Weekend in March 2004. Approximately 220 ONS Chapter Presidents and Presidents-elect from across the United States attended this meeting. Changes in the wording and format of the instrument/questionnaire were made prior to the distribution of the ONS Patient Outcomes Survey to the sample based upon the responses to the pilot study.
The ONS Patient Outcomes Survey was conducted from October 2004-December 1, 2004 using both web-based technology via Zoomerang.com and the US Mail. The ONS members with e-mail addresses were sent a bulk e-mail message including the content in a cover letter (approved by the University of Pittsburgh IRB) which explained the purpose of the study and their rights as a research subject. The e-mail message contained the Zoomerang® website address where the questionnaire was housed. Once the participants clicked on the website in the e-mail message, they were taken directly to the questionnaire. The ONS members without e-mail addresses received a hard copy of the cover letter and the questionnaire via U.S. Mail along with a postage-paid return envelop.
An incentive of an Oncology Nurse's Month pin was sent to the first ONS members to respond to the survey and submit their name/address separate from the survey (in order to maintain the anonymity of the subject) to the ONS Research Team. The first 200 nurses who responded to the survey using the Zoomerang® website and sent their name and address via e-mail to the ONS Research Team received a pin and the first 100 nurses who responded to the survey using the U.S. Mail and called the ONS Research Team using a toll-free number with their name and address received a pin.
The computer-based web surveys were entered directly by the study respondent using Zoomerang. The mail surveys included the same content as online survey. Once the mail surveys were received in the research office, an office staff member entered the questionnaire responses into the online questionnaire in Zoomerang. The data files from both the online surveys and the entered mail surveys were combined into a single data file. A variable included to specify route of response (e-mail or U.S. mail) was added so the U.S. mail respondents could be identified in the data file. Over eight in every ten (81.5%) who did reply, used the online survey option.
All of the data were screened for accuracy and completeness of entries and the open-ended responses were tabulated. As appropriate, open-ended responses were classified into categories for summary response tables. The data were initially tabulated for the complete sample. No weightings were applied, so the base frequencies represent the relative proportions among those who chose to respond. After generating the base descriptive frequencies from these data, we analyzed potential association with select characteristics by computing cross-tabulations using chi-square tests of significance. Associations are reported as significant if the p value for the chi-square test of the test of statistical independence was .05 or below.
Demographics of Respondents
The overall response rate for the survey was 13.6% (n = 1327/9778) with 12.5% (1083/8,678) response rate electronically and 22.2% (244/1,100) response rate via U.S. Mail. These response rates are very good response rates for a lengthy survey distributed to ONS general membership.
For Table 1: Respondents and ONS Membership Demographics (%), click here.
ONS members' perceptions of the impact, value, use and education needs of oncology nurses related to oncology nursing-sensitive patient outcomes
Overall, the vast majority of respondents indicated that they perceive that their nursing care has a positive impact on cancer patients' symptoms, functional status, psychological distress, and economic values.
Over 90% of the respondents indicated that they strongly agreed or agreed that they have a positive impact on pain (98.4%), nausea (98.2%), constipation (96.7%), diarrhea (93.6%) and mucositis (91.9%). The symptom receiving the lowest percentage of "strongly agree" or "agree" responses was anorexia (75.1%).
Over 90% of the respondents indicated that they strongly agreed or agreed that they have a positive impact on activities of daily living (90.9%). The functional status category receiving the lowest percentage of "strongly agree" or "agree" was role functioning (72.4%).
Over 90% of the respondents indicated that they strongly agreed or agreed that they have a positive impact on infections (96.9%) and extravasation incidents (90.0%). The safety category receiving the lowest percentage of "strongly agree" or "agree" was decubiti (72.5%).
Over 90% of the respondents indicated that they strongly agreed or agreed that they have a positive impact on anxiety (97.8%). The psychological distress category receiving the lowest percentage of "strongly agree" or "agree" was spiritual distress (77.0%).
The economic outcomes category received the lowest percentage of "strongly agree" or "agree" of all of the outcomes categories. The highest percentage of respondents indicated that they strongly agreed or agreed that they have a positive impact on emergency visits to the Emergency Department (64.5%). The lowest percentage of "strongly agree" or "agree" responses was for the category of cost per patient day (37.7%).
Over 90% of the respondents strongly agreed or agreed with the following statements regarding the value of nursing-sensitive patient outcomes:
- Patients/families are concerned about the outcomes of care (98.1%)
- It is important that nursing continuing education programs include information regarding nursing-sensitive patient outcomes (97.8%)
- It is important that staff development programs on nursing-sensitive patient outcomes be provided (97.2%)
- It is important that nursing education programs include information regarding nursing-sensitive patient outcomes (96.8%)
- I consider patient outcomes when I provide care (96.4%)
- It is important for nurses to document patient outcomes as a regular component of care (93.5%)
- Poor patient outcomes lead to increased cost of care (90.1%)
Over 90% of the respondents strongly disagreed or disagreed with the following statements regarding the value of nursing-sensitive patient outcomes:
- Concern about nursing-sensitive patient outcomes is not needed because the emphasis on health care is on cure and survival (94.3%)
- In general, nurses are not concerned about the outcomes of their care (91.4%)
Use-Assessment and Documentation
Over 80% of the respondents indicated that they find that patient outcomes measurement:
- Improves quality of care (86.8%)
- Helps with clinical decision-making (84.8%)
- Helps with patient assessment (80.1%)
73% of the respondents strongly agreed or agreed that they document nursing-sensitive patient outcomes in their practice and between 48.8% (psychological distress) and 67.9% (symptoms) strongly agreed or agreed that they use the same tool/instrument each time they document. Only 20.3% strongly agreed or agreed that they use the same tool/instrument to document cost outcomes.
The Numeric Pain Scale was reported as always used by 83.4% of the respondents (14.2% sometimes). Other specific tools reported used were:
- Brief Pain Inventory (38.5% always, 32.4% sometimes)
- Numeric Fatigue Scale (29.2% always, 33.1% sometimes)
- Index of Nausea, Vomiting, and Retching (19.4% always, 24.8% sometimes)
- Oral Assessment Guide (34.3% always, 31.7% sometimes)
The following tools/instruments were frequently reported as never used:
- SF 12 (93.2%)
- SF 36 (92.8%)
- CES-D (92.7%)
- Sloan Kettering Symptom Inventory (91.8%)
- Holland's Distress Thermometer (90.9%)
- Pittsburgh Sleep Quality Index (89.9%)
- MD Anderson Symptom Inventory (89.1%)
- Beck Depression Inventory (79.2%)
- Press-Ganey Satisfaction (65.1%)
Use-Standardized Nursing Languages
Nearly half of the respondents (33.8% - 50.3%) indicated that they were "not sure" if their institution utilized any of the listed Standardized Nursing Languages. 28.7% of the respondents reported that they used the North American Nursing Diagnosis Association Classification (NANDA).
Use-Contribution to Data Repositories
The largest percentage of the respondents reported that their practice setting contributed to the National Comprehensive Cancer Center Network - NCCN - Data Repository (18.4%). The majority of respondents responded that they were "not sure" (47.7% - 55%) if their practice setting contributed to any of the listed Data Repositories.
Over 92% of the respondents indicated that they strongly agreed or agreed that they needed more information about oncology nursing-sensitive patient outcomes in general. Needs for more information about specific outcome categories results were (% strongly agree/agree):
- Psychological distress outcomes (91.5%)
- Symptom management outcomes (89.7%)
- Functional status outcomes (89.7%)
- Safety outcomes (83.9%)
- Economic outcomes (82.1%)
Availability of information on nursing-sensitive patient outcomes
The largest percentage of respondents indicated that information about nursing-sensitive patient outcomes should be available on the ONS Website (formerly ONS Online) (92.0%). Other information sources included:
Clinical Journal of Oncology Nursing (81.3%)
Oncology Nursing Forum (78.6%)
- ONS Congress (70.9%)
- ONS Institutes of Learning (66.4%)
- Chapter meetings (65.6%)
ONS News (63.6%)
- ONS regional conferences (55.3%)
Similarities and differences in responses to the ONS Patient Outcomes Survey of Pennsylvania ONS members and a national sample of ONS members.
One of the questions examined for potential impact on patterns of the survey responses was whether or not the respondent worked in Pennsylvania. There were minimal significant differences between those who work in Pennsylvania and respondents from out of the state. The only significant differences were for two of the assessment measures reported in question number eight. Pennsylvania respondents were significantly more likely to report use of the Pittsburgh Sleep Quality Index, although the large majority of each group never use this assessment tool. Pennsylvania-based respondents were also significantly more likely to report the use of the Press-Ganey Satisfaction measure. Over half of the Pennsylvania-based respondents reported using this assessment 'always' or 'sometimes' in contrast to about one-third of the non-Pennsylvania respondents.
Similarities and differences in responses to the ONS Patient Outcomes Survey of ONS members with e-mail (Internet) access and ONS members without e-mail (Internet) access (do ONS members with e-mail access respond differently than ONS members without e-mail access?).
The large majority, 81.5 percent, of the survey respondents used the electronic submission option. About one in every six (18.5%) used the mail survey option. There were major differences across primary position in terms of the response method selected. Staff nurses were, by far, the most likely to respond using the mail option. Among this group, three in every ten (30.4%) used the mail survey option. No other primary position group had over 10 percent selecting the mail response option. Educators, Clinical Nurse Specialists, Clinical Trial Nurses, and Directors all had over 95 percent using the electronic survey to respond.
Placed in another perspective, Staff Nurses represented 49.1 percent of all survey respondents, but comprised 80.1 percent of those who selected the mail survey option.
Among the Staff Nurses, we examined the question of whether or not there were differences in the survey responses across methods. There were some patterns of significant differences across survey method for the Staff Nurses. For example, Staff Nurses who used the mail response method provided significantly more affirmative answers to five of the assessments of positive impact of nursing care on cancer patient outcomes (fatigue, insomnia, anorexia, neutropenia and hypersensitivity reactions). The mail response group of the Staff Nurses also give significantly more positive responses on two of the questions on the assessment and documentation of oncology nursing-sensitive patient outcomes (i.e., identified and used in my practice setting, use the same cost assessment outcomes each time).
The Staff Nurses who used the online survey, were more likely than the mail respondents to give affirmative responses to three of the questions on the importance and value of nursing-sensitive patient outcomes (i.e., patients/families concerned about outcomes, documenting nursing activities more important, nursing education programs should include nursing outcomes).
In summary, the use of electronic surveys among non-Staff Nurse populations would appear to provide good coverage with little risk of excluding segments of these groups. However, a substantial proportion of Staff Nurses selected to use the mail option in the fall of 2004 and, in selected areas of response, the mail survey respondents give answers that differed from the electronic responses. Thus, for surveys of Staff Nurses, some consideration should be given to the potential for coverage and response variation issues if using a single data collection mode.
Primary Inpatient/Outpatient work setting.
Although it was not an original aim of the survey (inpatient vs. outpatient), there were significant differences found when analyzing the responses by work setting. By analyzing the responses by work setting, we were able to find some significant differences. As oncology care is increasingly being performed in an outpatient or ambulatory setting, we felt it was important to assess for differences. This analysis provided information about perceptions of the respondents in different settings on oncology nursing-sensitive patient outcomes, but can also determine the direction for future research and education program planning.
About one in every three respondents (33.7%) worked in an inpatient setting (e.g., oncology specialty unit, medical surgical unit, intensive care unit). Fifty percent of the respondents reported working primarily in an outpatient setting (e.g., hospital based clinic, hospital based radiation center, free-standing radiation center, physicians office, home care, or hospice). The remainder (16.3%) were based in schools of nursing or other work settings.
An analysis of the pattern of responses to the core survey questions across inpatient/outpatient work setting shows a number of statistically significant contrasts. For example, on the first question related to nursing care impact on cancer patient symptoms, there were significant differences on eight of the ten symptoms. Interestingly, the differences were not all in the same direction. Respondents who worked primarily in inpatient settings provided significantly more affirmative (i.e., strongly agree or agree) answers to having positive impact on pain, fatigue and breathlessness. On the other hand, respondents who worked primarily in outpatient settings provided significantly more affirmative answers for the symptoms of nausea, constipation, anorexia, diarrhea, and mucositis.
In the questions on impact for functional status, the outpatient setting respondents provided significantly more affirmative responses for role functioning and nutritional status.
With respect to the questions on safety, inpatient-setting respondents were significantly more affirmative for falls and decubiti. Those working in outpatient setting reported more affirmative responses for hypersensitivity reactions.
On the economic outcomes questions, the inpatient setting respondents reported more affirmative answers for length of stay in hospital. In contrast, more affirmative responses for the outpatient setting respondents was evident in emergency visits, out-of-pocket costs, and patient co-payments.
When rating the importance or value of oncology nursing-sensitive patient outcomes, the respondents who worked primarily in inpatient settings were significantly higher in their affirmative responses for 7 of the 18 items (e.g., physician's practices too concerned with treatment protocols and drug trials, not possible to separajte nursing outcomes, standardized assessment tools, inclusion in nursing education programs, inclusion in continuing nursing education, staff development programs, and poor patient outcomes increasing cost of care. The outpatient site respondents were not more likely to give affirmative responses to any of these questions.
Summary of Key Findings
All of the aims of this survey were met. The demographics of the respondents were similar to the overall characteristics of ONS members except for a larger number of respondents from hospital-based clinics and medical oncology settings and more respondents with greater than 20 years of experience in nursing and in oncology nursing.
The respondents reported that they have a positive impact on cancer patients' symptoms, functional status, safety, psychological distress and economic values. The respondents value ONSPO, and want additional information about them. However, while the respondents indicated that they understood the importance of outcomes, the manner in which they measured and documented the outcomes varied. The Numeric Pain Scale was the most used tool, which is not surprising as both the American Pain Society and the Joint Commission and Association on Health Care have been encouraging this use since 2000. There were also inconsistencies in the nursing language used for documentation. While less than a third of the respondents used the North American Nursing Diagnosis Association Classification (NANDA), almost half of the respondents were not able to state what nursing language (if any) their employers used. This could be important for future research in nursing-sensitive patient outcomes, as it would be important to use the same language when comparing the impact of nursing interventions on outcomes patient tools. Nurses indicated that they desired additional education in the areas of psychological distress, functional status and safety outcomes, along with support in symptom management outcomes and economic outcomes.
Oncology nurses who reside in PA responded similarly to those who do not reside in PA. The only significant differences were with some of the measurement tools used (sleep and patient satisfaction). The similarity of responses of nurses who reside in PA to a national sample of nurses is important in analyzing the results of larger studies using PA nurses as a representative sample of nurses in the United States.
There were significant differences found in the response mode (email versus US mail), in role category (most of the nurses who responded by mail were staff nurses, whereas most non-staff nurses responded by e-mail) as well as values. Despite the differences found in use of e-mail, 92% of the respondents indicated that they felt that information about nursing-sensitive patient outcomes should be available on the ONS Website. This may indicate that while some nurses still prefer not to use e-mail for correspondence, they are comfortable accessing the ONS Website for information.
The results of this survey were important in providing ONS with information regarding the knowledge, perceptions and values of oncology nurses regarding oncology nursing-sensitive patient outcomes. Based on the results of this study, future research and education can focus on the needs of the respondents - namely, providing more information in the areas of psychological distress outcomes, symptom management outcomes, safety outcomes, and economic outcomes. Additional research regarding effective clinical and research measurement instruments and tools can lead to standardized measurement of outcomes. Finally, research can be done to determine the impact oncology nurses have on specific patient outcomes.
The results from this survey, combined with the results from other health services research projects regarding cancer patients' outcomes of care, can be used to design research to test the impact of specific oncology nursing interventions on oncology patients' outcomes. Oncology nursing interventions have been shown to reduce health risks related to cancer treatment in clinical studies. The results from this survey can be used to plan future large-scale studies of the delivery of health services (specifically oncology nursing interventions) to people with cancer. The mission of the Oncology Nursing Society is to promote excellence in oncology nursing and quality cancer care. One of the strategic goals of the Oncology Nursing Society is to generate public awareness of the specific contributions that oncology nurses and oncology nursing research make to patient outcomes. The results from this survey of oncology nurses will add to hospital level information on cancer patient outcomes currently being studied and provide direction for the Oncology Nursing Society's education and research efforts regarding the quality of care provided to people with cancer.