MHSIP Consumer Survey Technical Report Fiscal Year 2013
MHSIP Consumer Survey Technical Report Fiscal Year 2013 A Report from the Colorado Department of Human Services Office of Behavioral Health, Community Programs This report was prepared by: Adrienne Jones, M.A., Researcher, Data and Evaluation Unit, Community Programs Office of Behavioral Health
Table of Contents About this Report . 4 What is the MHSIP Survey? . 4 Survey Procedures . 4 Sample . 5 Survey Distribution . 5 Results . 5 Response Rate . 5 Respondent Demographics . 5 Race and Ethnicity . 5 Place of Residence . 5 Language . 6 Disability . 6 Employment . 6 Criminal Background . 6 Payor Source . 6 Health Services Utilization and Treatment Duration . 6 Consumer Input for Future Surveys . 6 FY2012 and 2013 Demographic Comparison . 7 MHSIP Respondents Compared to the CCAR Population. 8 Overall Domain Results . 8 Demographics and Domain Agreement . 10 Gender . 11 Age . 11 Ethnicity . 12 Race . 12 Sexual Orientation . 13 Disability . 13 Employment . 13 Language . 14 Medicaid Status . 14 Qualitative Comments . 14 Discussion and Implications . 16 Appendix A: MHSIP Survey . 18 FY2013 MHSIP Technical Report 15
Appendix B: Payor Status . 22 Appendix C: Response Rate by Agency . 23 Appendix D: Demographic Information of FY2013 MHSIP Respondents . 24 Appendix E: Comparison of Demographic Information of FY2013 MHSIP to FY2013 CCAR Respondents . 26 Appendix F: Percent Endorsement of MHSIP Domains by Item. 27 FY2013 MHSIP Technical Report 3
About this Report In 2012, the Colorado Office of Behavioral Health (OBH) conducted its sixteenth annual Mental Health Statistics Improvement Program (MHSIP) Consumer Survey with a focus on services provided in State Fiscal Year 2013 (July 1, 2012 - June 30, 2013). Consistent with national trends in performance measurement, OBH administers the MHSIP Consumer Survey to assess perceptions of public behavioral health services provided in Colorado. This report describes data collection, sample selection, and results of this year’s survey. OBH is committed to the inclusion of consumer participation at multiple levels of behavioral health services and perceives the MHSIP survey as one way of meeting this ongoing goal. It is important to note that the MHSIP survey has been developed at a national level in part to promote data standards that allow for valid results to better inform policy and decisions (for a full description of MHSIP and the survey’s underlying values, please visit: http://www.nri-inc.org/projects/SDICC/tech assist.cfm. MHSIP work groups include consumers and families with the seminal aim of such groups being the promotion of consumer-oriented services through data. OBH has a vested interest in promoting these values in Colorado as the state moves toward a recoveryoriented behavioral health system. Continuing the national-state MHSIP partnership is key to this endeavor. As evidence of the weight that OBH has placed on the promotion of consumer-driven services, it is notable that the MHSIP has been incorporated into multiple levels of operations, including a federal grant application and statewide mental health center contracts. The MHSIP survey continues to provide an excellent opportunity for OBH to partner on both national and statewide levels to shape future services through data. Thank you to all who assisted in the data collection of the MHSIP survey. Agency collaboration is instrumental to the success of the survey and OBH acknowledges and appreciates the hard work of the mental health centers and clinics in this process. OBH would also like to extend a special thanks to Angie Lawson, Ph.D. for her help on this report. What is the MHSIP Survey? The MHSIP Consumer Survey consists of 36 items, each answered using a 5-point Likert scale ranging from one (strongly agree) to five (strongly disagree; see Appendix A). Standardized at a national level, the survey comprises the five following domains: Access: six items that assess perceptions about service accessibility Quality/Appropriateness: nine items that assess perceptions of quality and appropriateness Outcomes: eight items that assess perceptions of outcomes as a result of services Participation: two items that assess perceptions of consumer involvement in treatment General Satisfaction: three items that assess satisfaction with services received Additionally, one item assesses perceived provider sensitivity to cultural/ethnic backgrounds of consumers. The questionnaire also contains items pertaining to demographic information (e.g., age, gender, and ethnicity). Two open-ended questions are also included in order to gather opinions about the most and least preferred aspects of services received. OBH distributes the MHSIP Consumer Survey in both English and Spanish. Survey Procedures A convenience sample has been used since 2008 where surveys were given directly to consumers when they arrived for their appointment, meaning that these consumers were currently receiving services. In previous survey administrations, the surveys were mailed to current and discharged consumers (prior to FY2009). Additionally, consumers who chose to complete the MHSIP survey were eligible to enter a drawing to win a 10 gift card for a local grocery or convenient stores. All consumers were included in the survey regardless of payor source. This was different from previous years that included only the consumers who were classified as indigent or receiving Medicaid. For more information on Payor Source, please see Appendix B. FY2013 MHSIP Technical Report 4
Sample The Office used a convenience sampling method whereby each of the 17 community mental health centers and the two specialty clinics, Asian Pacific and Servicios de la Raza, were provided with surveys to hand out to consumers who were receiving services during a three week period conducted between October 8th and October 26th 2012. Consumers who were attending a first appointment or an intake were excluded from the sample. Survey Distribution OBH contracted with the State of Colorado Central Services, Integrated Document Solutions (IDS) department to prepare, mail, and receive surveys as well as enter data for the FY2013 survey period. IDS mailed a pre-determined number of MHSIP packets (including a cover letter, survey, and a lottery ticket) to each of the 17 community mental health centers and the two specialty clinics. During the three-week data collection period, consumers were offered the opportunity to complete the MHSIP survey and a lottery ticket for entering a gift card drawing. Consumers could choose to mail the survey directly to IDS in a postage-paid return envelope or could drop (completed and refused) surveys in a secure box located at the center. At the end of the data collection period, centers shipped all surveys (completed and refused) to IDS where they were sorted and processed. Data from the completed surveys were then entered and forwarded to OBH by IDS. Results The unit of analysis for this report is at a state level. Although OBH previously computed domain scores at the agency level, this approach was stopped because it undermined OBH’s goal to foster a collaborative and learning environment amongst Colorado’s public mental health system. Rather, scores are computed at the state level and individual agencies are given the data upon request for their specific agency for further analysis. Response Rate Response rate was calculated by taking the number of surveys completed from each agency divided by the number of consumers scheduled for appointments. When calculating response rate in this way, it is assumed that every consumer was offered the survey and either declined or accepted. The Office received a total of 3,338 completed or partially completed surveys. Soon after the survey administration, agencies were asked for the number of consumers that were seen during the survey period. All agencies responded and 21,240 adult consumers were reportedly seen during the 3-week survey period, representing a 15.7% return rate (see Appendix C). This percentage is higher than FY2012, which had a response rate of 11.4%. This year’s survey process was not significantly different from the three previous years, and the number of surveys completed this year was higher than numbers in recent years (2,396; 2,327; and 2,642 respectively). Respondent Demographics The majority of the MHSIP respondents who reported gender were female (63.2%) and 35.9% were male. Respondents were generally middle aged with 29.3% between 31-45 years old and another 25.4% between 46-64 years old. Regarding sexual orientation, respondents were asked to self-identify their sexual orientation. Of those who chose a sexual orientation, the majority identified as “heterosexual” (75.4%) followed by “bisexual” (4.4%), “other” (4.1%), and “lesbian or gay” (3.2%). A small percentage (1.3%) chose more than one sexual orientation. Many respondents (11.5%) either left the item blank or marked “Prefer not to Answer.” Race and Ethnicity: Following national guidelines, race and ethnicity were separated into two questions on the survey. Hispanic/Latino(a) was the sole choice for ethnicity and 21.3% of respondents endorsed this item. However, 30.9% of the responses were missing or marked “Prefer not to Answer.” Race had the following choices: American Indian/Alaska Native, White/Caucasian, Black/African American, Native Hawaiian/Pacific Islander, Asian, Other, and Prefer not to Answer. If a respondent chose more than one race, their racial identification was coded as Multiracial. Most respondents identified with only one racial group (87.6%). The majority of respondents identified as White/Caucasian (73.5 followed by Multiracial (6.5%), Other (5.1%), African American/Black (4.3%) and American Indian/Alaska Native (3.2%); (see Appendix D for all responses). Approximately 17.3% left this item blank or chose “Prefer not to Answer.” FY2013 MHSIP Technical Report 5
Place of Residence: With respect to place of residence, 48.9% of respondents indicated that they lived within 5 miles of the mental health center, 28.1% lived 6-10 miles away, 15.9% lived 11 to 20 miles away, and 7.1% lived more than 20 miles away. Relationship Status and Military Service are presented in Appendix D. Language: Regarding language fluency, 13.7% of respondents were bi- or multi-lingual while the majority of respondents spoke one language fluently (85.4%). Of those respondents that spoke one language, most often the language was English (83.3%) followed by Spanish (1.7%). For more languages spoken information, see Appendix D. Disability: Sixteen percent of the respondents in this survey left these items blank or chose “Prefer not to Answer,” which represents a substantial change from last year’s respondents, of which over half left the item blank or chose “Prefer not to Answer.” Of those who chose other responses, (54.3%) reported having at least one type of disability (excluding mental health, although those who chose “Other” may have filled in a mental health disability). Almost one-fourth of respondents (21.2%) identified as having multiple disabilities. Of those with one disability endorsed, the highest reported were a physical disability (12.2%), followed by learning disability (6.6%). Regardless of number of disabilities endorsed, the highest reported were physical disability (23.8%) and learning disability (17.8%). Employment: Regarding employment, 69.7% reported not having worked at a paid job in the three months prior to the survey; however, 25.5% of the sample indicated having volunteered in this time frame. Criminal Background: FY2013, the survey asked about criminal background (i.e., arrests) for respondents who had received services for more than one year and for those who had received services for less than one year. For respondents who had received services for more than one year, 10.4% of survey respondents reported having been arrested in the past 12 months with a slightly lower proportion (9.0%) indicating having been arrested in the 12 months prior to that time frame. For those who had received services for less than one year, 8.4% reported having been arrested since beginning mental health services. A slightly higher proportion (16.9%) were arrested in the 12 months prior to beginning mental health services. Please note that a significant percentage (43.9%, 45.2%, 34.4%, and 35.0%, respectively) did not answer these questions at all. Payor Source: On this year’s survey, there were three ways of collecting data regarding respondents’ payor source. There were three specific items on the survey: “Do you currently receive Medicaid?” with a yes/no answer; “What other form of insurance do you have?” with multiple options; and “What form of payment best describes your payment plan for services here?” with multiple options. In addition, agency staff were asked to mark up to three payor sources of the respondent on the survey when handing it to them. According to respondents answering the survey question, 59.7% were currently receiving Medicaid at the time of survey completion (with only 7.1% of respondents missing data on this item). The next highest endorsed option was “no insurance,” with 35.7% of the respondents endorsing this choice. The option of “Medicaid/Medicare co-pay” was the most frequently chosen response (48.2%) to the payment plan for services, followed by “no payment” (14.9%). The agency data are not seen as an accurate representation of payor source this year, as 42.2% were not filled out in the designated spot by staff on the surveys. Health Services Utilization and Treatment Duration: Among FY2013 MHSIP respondents, 66.2% indicated having seen a physician or nurse for a health check-up, physical exam, or for an illness during the past year and an additional 10.7% indicated having seen a physician or nurse in an emergency room visit. Respondents were asked to report the number of sessions they had attended in the last six months ranging from 0 to 26 or more sessions. Most frequently, respondents had attended 1-5 sessions (33.3%) or 6-11 sessions (24.2%). However, a little over 39.4% received 12 or more sessions, with 17.0% of MHSIP respondents attended 12-18 sessions, 9.7% attended 19-25 sessions, and 12.7% attended 26 or more sessions. Additionally, 18.0% of respondents reported that they are required by someone else (e.g., social services, court-ordered) to attend sessions. Lastly, a majority of respondents (68.0%) reported that they were receiving medication treatment through their agency at the time of survey completion. Consumer Input for Future Surveys: On this year’s survey, there was a question asking MHSIP respondents for input into future surveys. Specifically, the question was, “In the future, would you like to complete this survey FY2013 MHSIP Technical Report 6
online?” Approximately 60 % of respondents reported they would like to do so, and approximately 40% reported that they would not like to do so. For FY2014, an online survey option will be available to consumers. FY2012 and 2013 Demographic Comparison: In terms of comparing MHSIP respondent demographics from year to year, the demographics of FY2012 are similar to FY2013 demonstrating that the populations are highly comparable. See Figures 1 and 2 for demographic data. Figure 1. Age of respondents by percentage, FY2012-FY213 MHSIP Age Comparison FY2012 and FY2013 40% 30% FY2012 20% FY2013 10% 0% 18-20 21-30 31-45 46-64 65-74 75 Figure 2. Race of respondents by percentage, FY2012-FY2013 MHSIP Race Comparison FY2012 and FY2013 80% 70% 60% 50% 40% 30% 20% 10% 0% FY2012 FY2013 FY2013 MHSIP Technical Report 7
MHSIP Respondents Compared to the CCAR Population The MHSIP respondents’ demographics were examined in comparison to demographics for the CCAR population. The CCAR measure is a more complete and thus more representative sample of people receiving mental health services within the state. Therefore, the samples were compared to explore whether the MHSIP sample is representative of this larger group. Statistical comparisons were not made as the sample sizes vary greatly in size and the instruments measure demographics slightly differently and are completed by different individuals (the MHSIP is self-report and the CCAR is clinician report). Instead, the comparison was an overall examination of trends of demographic similarity. The two samples were compared on gender, age, race, ethnicity, marital status, and paid employment. In general, the two samples are demographically similar: more women receiving services than men, single/never married as the primary marital status, and adults (21-65) as the primary age group served. Respondents identifying as Hispanic/Latino(a) in both samples was similar with 21.3% on the MHSIP and 20.0% on the CCAR. The demographic variable that differed the most (percentage-wise) was paid employment. For paid employment, MHSIP respondents reported a higher level of employment. This may be a result of the CCAR having many more choices regarding employment status than the MHSIP, which has a yes/no question only. It may also be that clinicians are not always aware of a consumer’s work status and that self-report impacted this variable. Please see Appendix E for CCAR and MHSIP demographic data. Overall Domain Results OBH computes domain scores reflecting the percentage of agreement versus disagreement for the State of Colorado. Agreement is defined as a mean that ranges from 1 to 2.49, whereas disagreement is defined as a mean that ranges from 2.50 to 5. Respondents who did not answer at least 2/3 of domain items did not receive a domain score. This method of computation follows national recommendations. Figure 3 represents the MHSIP questions by each domain. FY2013 MHSIP Technical Report 8
Figure 3. MHSIP Domain Items Satisfaction I liked the services that I receieved here. If I had other choices, I would still get services from this agency. I would recommend this agency to a friend or family member. Access The location of services was convenient. Staff were willing to see me as often as I felt it was necessary. Staff returned my calls within 24 hours. Services were available at times that were good for me. I was able to get the services I thought I needed. I was able to see a psychiatrist when I wanted to. Quality Staff here believe I can grow, change and recover. I felt free to complain. Staff told me what side-effects to watch for. Staff respected my wishes about who is, and is not, to be given information about my treatment. Staff were sensitve to my cultural/ethnic background. Staff helped me obtain information so that I could take charge of managing my illness. I was encouraged to use consumer-run programs. Participation I felt comfortable asking questions about my treatment and medication. I, not staff, decided my treatment goals. Outcome I deal more effectively with daily problems. I am better able to control my life. I am better able to deal with crises. I am getting along better with my family. I do better in social situations. I do better in school and/or work. My housing situation has improved. My symptoms are not bothering me as much. FY2013 MHSIP Technical Report 9
Table 1 presents summary results in percentages with confidence intervals (95%) for the total scores for FY2013, as well as for 2010/2011 and 2012. Looking at the trends over time, FY2013 levels of agreement remained relatively stable for all domains compared to FY2010/2011 and 2012. Please refer to Appendix F where percentages of endorsement for the 5-point Likert scale are presented by item within each domain. Table 1. Valid Domain Percent Agreement by Fiscal Year Access Fiscal Year 2010/2011 (95%CI) (n) 2012 (95% CI) (n) 2013 (95%CI) (n) 84.9 (83.4-86.4) (1951) 83.1 (81.6-84.6) (1973) 84.8 (83.6-86.1) (3301) Quality/ Appropriateness 89.6 (88.3-90.9) (2016) 88.7 (87.5-89.9) (2069) 90.7 (89.7-91.7) (3246) Outcomes 66.8 (64.8-68.8) (1467) 66.1 (64.2-68.0) (1506) 64.7 (63.1-66.4) (3180) Participation 79.8 (78.1-81.5) (1759) 79.6 (78.0-81.2) (1812) 80.8 (79.5-82.2) (3178) General Satisfaction 90.6 (89.4-91.8) (2098) 90.3 (89.2-91.4) (2150) 90.1 (89.1-91.1) (3315) Due to consistent methodology, it is possible to examine trends in domain scores from year to year. Figure 4 demonstrates that the domain scores for these three years are consistent and follow the same overall trend in percent agreement. Figure 4. MHSIP Domain Scores, FY2010/11 through FY2013 MHSIP Domain Score: FY2010/11 - FY2013 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% FY2010/2011 FY2012 FY2013 Demographics and Domain Agreement Analysis of variance (ANOVA) tests were conducted examining relations between domain agreement and the demographic and other variables recorded on the MHSIP. To help correct for the high number of statistical tests run, a conservative approach was used for the interpretation of significance (p .001). Specifically, the variables of gender, age group, ethnicity, race, employment status, sexual orientation, relationship status, disability status, language, and payor source were examined with relation to domain agreement. Due to the large differences in FY2013 MHSIP Technical Report 10
sample size between “Prefer Not to Answer” and the other response choices for some of the questions, the demographic variables were analyzed without this response choice. Gender ANOVA results found that gender had a statistically significant impact on Access [F(4, 3187) 4.75, p .001], Quality/Appropriateness [F(4, 3154) 5.67, p .000], and Participation [F(4, 3194) 5.16, p .000]. Table 2. Valid Percent Agreement by Gender Gender (n) Access Quality/Appropriateness Outcomes Participation Woman (1,946) Man (1,096) Transgender ( 10) Other ( 10) 85 86 78 80 89 87 79 80 62 67 63 80 83 79 78 80 General Satisfaction 91 90 78 100 Note. The reported n of each gender category reflects the smallest number of total respondents on any one domain. The number of respondents across domains fluctuated by a small amount. Age ANOVA results did not find any overall significant difference between scores on the domains by age group using a significance level of p .001. It is important to note that, when respondents were divided by age group, the sample sizes of each group became relatively small and may have influenced results. Table 3. Valid Percent Agreement by Age Group Age Group (n) 18-20 (84) 21-30 (450) 31-45 (934) 46-64 (810) 65-74 (67) 75 ( 10) Access Quality/ Appropriateness Outcomes Participation General Satisfaction 86 85 83 86 69 100 88 88 88 87 90 94 68 67 62 64 76 60 73 82 83 80 87 94 90 88 90 90 95 100 Note. The reported n of each age category reflects the smallest number of total respondents on any one domain. The number of respondents across domains fluctuated by a small amount. FY2013 MHSIP Technical Report 11
Ethnicity A significant difference was found between scores on all domains by ethnicity using a significance level of p .01 A t-test was used for this analysis because there are only two response choices (See Table 4 for percent agreement across domains by Ethnicity groups). Significance score are as follows: Access: t(2291) 3.15, p .002 Quality/Appropriateness: t(2294) 5.62, p .000 Outcomes: t(2260) -22.6, p .000 Participation: t(2294) 5.62, p .000 General Satisfaction: t(2296) 12.13, p .000 Table 4. Valid Percent Agreement by Ethnicity Ethnicity (n) Access Quality/Appropriateness Outcomes Participation 88 85 91 90 63 63 81 84 Hispanic (538) Non-Hispanic (1698) General Satisfaction 92 89 Note. The reported n of each ethnicity category reflects the smallest number of total respondents on any one domain. The number of respondents across domains fluctuated by a small amount. Race Statistical significance in the race category was only found within the Access domain [F (6, 2733) 3.69, p .001]. See Table 5 for percent agreement across domains by Race groups. Table 5. Valid Percent Agreement by Race Race (n) American Indian/Alaska Native (91) Asian (25) Black/African American (124) White/Caucasian (2060) Native Hawaiian/Pacific Islander ( 10) Multiracial (190) Other (151) Access Quality/ Appropriateness Outcomes Participation General Satisfaction 87 97 89 94 77 68 79 89 94 91 84 86 88 89 62 64 80 82 89 91 80 78 84 80 83 87 80 69 60 80 80 79 80 86 92 Note. The reported n of each race category reflects the smallest number of total respondents on any one domain. The number of respondents across domains fluctuated by a small amount. FY2013 MHSIP Technical Report 12
Sexual Orientation ANOVA results found significance between sexual orientation and Access [F(4, 2649) 4.41, p .001] as well as sexual orientation and Outcome [F (4, 2616) 4.78 p .001]. See Table 6 for the percent agreement of all domains by sexual orientation. Table 6. Valid Percent Agreement by Sexual Orientation Sexual Orientation (n) Bisexual (94) Heterosexual (2211) Lesbian/Gay (93) Multiple Responses (36) Other (120) Access Quality/Appropriateness Outcomes Participation General Satisfaction 76 86 88 83 89 93 50 65 62 74 83 86 84 91 93 82 87 87 92 64 65 87 79 87 96 Note. The reported n of each sexual orientation category reflects the smallest number of total respondents on any one domain. The number of respondents across domains fluctuated by a small amount. Disability ANOVA results indicate significant differences between scores on the Outcome domain by disability [F (8, 2721)11.21 p .000]. See Table 7 comparisons of percent agreement for all domains by disability. Table 7. Valid Percent Agreement by Disability Disability (n) Blind/Partially Sighted (30) Deaf/Hard of Hearing (47) Developmental (85) Learning (192) None (770) Physical (363) TBI (56) Multiple Disabilities (615) Other (185) Access Quality/Appropriateness Outcomes Participation General Satisfaction 87 98 85 89 87 86 79 86 92 85 89 90 89 86 67 57 65 63 73 64 67 70 84 78 85 84 82 79 90 92 87 92 92 89 77 82 82 87 88 54 60 80 81 90 89 Note. The
1, 2012 - June 30, 2013). Consistent with national trends in performance measurement, OBH administers the MHSIP Consumer Survey to assess perceptions of public behavioral health services provided in Colorado. This report describes data collection, sample selection, and results of this year's survey. OBH is committed to the inclusion of
Improvement Program (MHSIP) Consumer Survey with a focus on services provided in State Fiscal Year 2012 (July 1, 2011-June 30, 2012). Consistent with national trends in performance measurement, DBH administers the MHSIP Consumer Survey to assess perceptions of public behavioral health services provided in Colorado. This report
Health Services (CMHS) t o conduct a survey of individuals served perceptions of the mental health care they have received from the public community mental health system. The survey tool used is the Mental Health Statistical Improvement Project Survey (MHSIP) for Adults. The results from this survey tool are reported to CMHS for the Mental Health
CONSUMER SURVEY 2014. FROM THE MHSIP . This survey was sponsored by the National Center for Mental Health Services. B o b b i R e n n e r , P h . D . , Q u a l i t y A s s u r a n c e M a n a g e r. 2 . . The MHSIP survey included a few questions regarding health risks. The table below shows the percent of survey respondents answering .
The MHSIP Consumer Survey consists of 36 items, each answered using a 5-point Likert scale ranging from one (strongly agree) to five (strongly disagree). These questions create seven domains, which are used to measure different aspects of customer satisfaction with public behavioral health services. Figure 1 details the specific
Program Survey (MHSIP) for Adults and the Youth Services Survey for Families (YSS-F). The MHSIP survey presents statements about services within seven domains and asks respondents to state to what degree they agree or disagree with them. The domains and a sample statement from each domain are shown in Table 1. Table 1.
MHSIP Quality Report Workgroup with support from the Evaluation Center@HSRI, under the direction of Dow A. Wieman, Ph.D., Senior Research Associate, Human Services Research Institute, Cambridge, MA with . Determine Method of Consumer Survey Administration.16 Step 4. Decide on Data Collectors/Interviewers.20 Step 5. Establish .
2 Kentucky MHSIP 2011 Appropriate Use of These Survey Results The results of this survey are intended to express the aggregated data of respondents' perception of care. This booklet represents the perceptions of 5.35 % sample of CMHC clients who are actively receiving outpatient services during the time of the survey implementation.
4.3 STAGES OF SOCIAL WORK GROUP FORMATION There are a number of stages or phases in formation of a social work group. Ken Heap (1985) discussed these as group formation and planning; the first meetings; the working phase; use of activities and action; and the termination of the Group. According to Douglas (1979) there are five stages viz., conceptualisation, creation, operation, termination .
