Sri Irianti, Puguh Prasetyoputra and Focusing on the Importance of Segregation at Source and Color-Coded Collection System.

Transcription

1 Sri Irianti, Puguh Prasetyoputra and Sunil Herat, Determinants of Hospital Waste Management ISSN in Indonesia: V o l u m e 8, N u m b e r 2 : , J u n e, T2013 Department of Environmental Engineering S e p u l u h N o p e m b e r I n s t i t u t e o f T e c h n o l o g y, S u r a b a y a & Indonesian Society of Sanitary and Environmental Engineers, Jakarta International peer-reviewed journal O p e n A c c e s s h t t p : / / w w w. t r i s a n i t a. o r g / j a s e s This research paper is licensed under the Creative Commons Attribution 3.0 Unported License, which permits unrestricted use, distribution, and reproduction in any medium, provided the originall work is properly cited. DETERMINANTS OF HOSPITAL WASTE MANAGEMENT IN INDONESIA: FOCUSING ON THE IMPORTANCE OF SEGREGATION AT SOURCE AND COLOR-CODED COLLECTION SYSTEM. SRI IRIANTI 1 *, PUGUH PRASETYOPUTRA 1 and SUNIL HERAT 1 National Institute of Health Research and Development, Indonesian Ministry of Health. 2 Griffith School of Engineering, Griffith University, Queensland, Australia. *Corresponding Author: Phone: ; Fax: ; irianti@litbang.depkes.go.id Received: 15 th March 2013; Revised: 4 th June 2013; Accepted: 5 th June 2013 Abstract: Majority of hospitals in Indonesia do not manage their wastes, properly, to minimize their risks by ignoring segregation at source and color-coded collection system. Consequently, their wastes are unsafe for patients, healthcare workers, waste handlers and the generall population since they can spread health-acquired infections (HAI). A cross-sectional study was carried out in 2010 to determine factors of current hospital waste management (HCW) of sample hospitals employing a structured questionnaire in order to develop sustainable waste management. There were two models of statistical analyses for waste segregation and color-coded system using multivariate logistic regression. The study revealed that determinants of waste segregation were availability of budget, central policy, management policy, and availability of standard operating procedures (SOPs) in hospital wards. The overall model of segregation is significant (p<0.001) with R 2 = Whereas, the determinants of color-coded system were the availability of waste management unit, the room SOPs, hospital classes and location (p<0.001 and R 2 =0.3789).. The study concluded that all the significant variables were important components for establishing sustainable HCW emphasizing waste segregation at source and color-coded system to comply with relevant regulations. Keywords: Medical waste, hospital, multivariate analysis, sustainable management T 2 INTRODUCTION In line with population growth, changes in patterns of diseases, and health sector decentralization since 2000, the number of hospitals in Indonesia is evidently increasing. In 2006, there were 1,012 general hospitals, with 118,504 beds, across 33 provinces, comprising those that are owned by the Ministry of Health, local governments at provincial and district levels, the military, and the private sector [1]. The number of general hospitals annually increased to 1,299 in 2010 [1]. From that number, 593 general hospitals (45.65%) were owned 135

2 by private companies and the remaining (54.35%) were owned by the governments, with the total beds being 22,860, and 143,428, respectively. Hospitals, as part of healthcare establishments, despite delivering curative health services for general public also generate wastes as by-products of their activities. Although the proportion of medical waste that falls into the category of hazardous waste is only 15-25%, it certainly poses a great risk, since it can spread HAI including blood-borne illnesses, like hepatitis B, hepatitis C, and HIV/AIDS [2]. Prüss et al [2] provided data from Japan, regarding estimated risk of blood-borne diseases after hypodermic needle puncture. The risk of infection of HIV is 0.3%, and the risk of infection of viral hepatitis B and hepatitis C is 3% and 3-5%, respectively. In addition to the sharps related injury, there were no reliable data from Indonesian hospitals on occupationally-acquired sharps injuries and musculo-cutaneous exposure. Sasimartoyo [3] revealed that health care workers were reluctant to answer whether they have experienced puncture by used needles. He also failed to find any recorded data concerning sharp-related injuries from Indonesian health care institutions, including hospitals which are assumed to be better source of such information. In contrast, researchers in a 6-year retrospective study conducted in a teaching hospital in Australia [4] found that among health care workers, medical staff experienced the highest proportion of sharp injuries (10.4%), and hollow-bore needles were implicated in 51.7% of all percutaneous injuries. Most incidents occurred during sharp use (40.4%), or after use but before disposal (27.1%). Nursing staff experienced 68.5% of reported mucocutaneous exposure. Many such exposures occurred when gloves were not used. These findings are certainly evident of the importance of PPE to reduce waste-related injuries among health care workers. Health care waste generation is influenced by such factors as: the kind of hospital, number of infectious disease beds, total number of beds, and number of outpatients per day [5], [6]. Types of wards and health care services also influence the amounts and characteristics of waste generated from each ward. Therefore, a hospital s waste generation can be counted by identifying each type of waste, and weighing it to determine its type and volume or weight. In a health care institution with no waste management system, the wastes from each ward is usually collected and contained together with general waste. Until recently, there was no reliable data of the amount of HCW generated per bed from general hospitals, since there was no comprehensive study about HCWM, with waste audits and LCA [7]. The generation of HCW was also examined in 100 hospitals in Java and Bali islands, finding approximately, 3.2 kg.bed -1.day -1 of solid HCW [8]. The proportions of solid general and medical wastes were 76.8% and 23.2%, respectively. This data indicates that there was no segregation at source, since the fraction of medical waste was still high. However, there was no explanation about waste generation from different classes and types of hospitals. When appropriate waste management systems are instituted in a hospital, segregation at source is practiced, and medical and general wastes are separated, according to the relevant regulation. In more advanced HWM practices, sharps wastes and other infectious wastes, such as soiled bandages, human tissues, and laboratory wastes are also separated, depending on the types of treatment technologies. Diaz, et al [9] and Mbongwe et al [10] point out that a full understanding of characteristics of medical wastes will lead to better choices of appropriate technologies. Similarly, Tudor [11] highlighted the importance of medical waste definition and standardization of measurement units to establish sustainable healthcare waste management (HCWM), which is, currently limited. According to the findings of Sasimartoyo [3], only 55.3% of hospitals studied, implement segregation at source. Hence, segregation practice at source of waste streams should be 136

3 introduced to all health care personnel, and it will be the main phase of the waste management hierarchy (WMH). Segregation practice will be sustainable, if there are sufficient skilled personnel, infrastructure and facilities, like trained operators, color-coded bins, special trolleys, color coded plastics, SOPs, and adequate storages. Moreover, the segregation and collection based on color coded should be followed by treatment and disposal units, using appropriate methods and technologies with regard to the WMH. Therefore, the main issues of establishing safe HCWM are practicing appropriate waste segregation at point of generation and containment so that relevant method of waste treatment can be determined effectively. This paper will present determinants of sustainable HCWM with regard to the importance of appropriate waste segregation and containment which will be useful for implementing safe HCWM in Indonesia. MATERIALS AND METHODS The study design was cross-sectional with the sample size of 237 government hospitals which was equal to 50% of total government hospitals in 2009 and they were determined by simple random sampling technique. The questions regarding bed occupancy ratio, numbers of inpatients and outpatients, classes of hospitals, waste generation, segregation, containment, storage and treatment and other related variables of HCWM were collected using a structure questionnaire mailed to all population of hospitals. Moreover, the main content of the questionnaire was adopted from World Health Organization (WHO) questionnaires for HCWM rapid assessment [12], so that the internal validation was already tested. Ethical clearance was obtained from Health Ethic Committee of National Institute of Health Research and Development, Indonesian Ministry of Health (MoH) and approvals and informed consents were from sample hospitals. The data were processed and analyzed using IBM SPSS version 20 to obtain descriptive and inferential data based on multivariate logistic regression [13] and current knowledge of HCWM. There are two statistical models for segregation at source and color-coded system. The list of variables can be seen in Table 1 and the logistic regression models containing equations of variable involved are in Table 2. Table 1: List of explanatory variables used in both logistic regression models No. Label of variable Name of variable Coding 1. Availability of routine budget DWMBUDGET 1 = Yes; 0 = No 2. Existence of sanitation unit DWMUNIT 1 = Yes; 0 = No 3. Availability of HCWM plans DWMPLAN 1 = Yes; 0 = No 4. Availability of policy from Central DCENPOLICY 1 = Yes; 0 = No Government 5. Existence of policy on HCWM by DMANPOLICY 1 = Yes; 0 = No Hospital Manager 6. Existence of hospital guideline on DHOSPGUIDE 1 = Yes; 0 = No HCWM implementation 7. SOP in the room on segregation DLOCATION 1 = Yes; 0 = No 137

4 Table 2: Logistic regression models Model 1: Waste Segregation Model SEGREGATIONi = β1 + β2dwmbudgeti + β3dwmuniti + β4dwmplani + β5dcenpolicyi + β6dmanpolicyi + β7dhospguidei + β8droomsopi + β9dhospclassi + β10dlocationi + ei Logistic Regression Model 1 Model 2: Color Coding Model COLCODINGi = β1 + β2dwmbudgeti + β3dwmuniti + β4dwmplani + β5dcenpolicyi + β6dmanpolicyi + β7dhospguidei + β8droomsopi + β9dhospclassi + β10dlocationi + ei Logistic Regression Model 2 In order to be able to generalize the outcomes of inferential statistics, it is necessary to conduct a test to ensure that the models used do not suffer from severe multicollinearity. The assessment of multicollinearity in logistic regression models is similar to that in the ordinary multiple regression models. Table 3 shows a summary of the collinearity statistics from both, Waste Segregation, and Color Coding, models. In both, there are no values of VIF that exceed 10; there are no values of TOL below 0.20; and the average VIFs are both not very far from 1. In conclusion, therefore, both models can be assumed to be free from severe multicollinearity. Table 3: Collinearity statistics from Waste Segregation Model and Color Coding Model Variable TOL VIF Variable TOL VIF WMBUDGET HOSPGUIDE WMUNIT ROOMSOP WMPLAN HOSPCLASS CENPOLICY LOCATION MANPOLICY Note: TOL = Tolerance; VIF = variance-inflating factor. RESULTS AND DISCUSSION 138 Average VIF = 1.29 Solid waste streams and generation Concerning solid waste streams and generation, 218 out of 327 sample hospitals answered the question about solid waste production. There are variations in waste streams from hospitals, depending on their class, since the classes determine the types of medical services, types of specialties, and the numbers of beds. Class D hospitals, which are the lowest class, typically

5 have four types of specialist services, namely, pediatrics, obstetrics & gynecology, internist, and surgery. Waste streams can also be determined by disease patterns and treatment methods, and the length of stay (LOS). In terms of waste generation, hospital produced wastes include, general waste with characteristics similar to municipal waste, and medical waste, comprising infectious, chemical, pathological, radioactive, cytotoxic and other hazardous wastes. The weights were already adjusted to the occupied beds, based on the bed occupancy ratio (BOR). Hospitals generating kg.bed -1 day -1 (25.30%) formed the highest proportion, while a single hospital producing more that 6 kg.bed -1 day -1 was the lowest percentage. Overall, the generation of general waste ranged from 0.01 kg.bed -1.day -1 to more than 6.50 kg.bed -1.day -1. In comparison, the general waste generation in Indonesian hospitals is similar to the amount of general waste from Brazil (2.675 kg.bed -1.day -1 ) [14], Taiwan ( kg.bed -1.day -1 ) [6] and Iran (2.439 kg.bed -1 day -1 ) [15]. Although all hospitals generate medical wastes, only a small proportion of them did not weigh their medical wastes. From 237 hospitals, 221 (93, 20%) hospitals weighed their medical wastes, and only a small proportion of hospitals did not calculate their wastes. The highest proportion was hospitals that produced kg.bed -1.day -1 (39.80%) of medical wastes, followed by 53 (24.00%) hospitals that produced kg.bed -1.day -1. Only two hospitals generated more than 1.60 kg.bed -1.day -1 of medical wastes. The daily average of medical waste generation per bed was kg. In comparison, Brazil produced medical waste of kg.bed -1.day -1 [14], Taiwan generated kg.bed -1.day -1 of medical waste [6], and Iran produced kg.bed -1.day -1 [15]. Using this number to determine the proportion of hospitals, 137 (57.81%) hospitals produced below the daily average, weight of medical wastes. These numbers are very useful for planning waste management, including numbers of containers and plastic bags, storage, personnel, capacity of treatment technology for onsite treatment, needed. Solid waste segregation, collection and containment Waste segregation, collection and containment at source are the most important stages of WMH, when waste generation cannot be avoided. It is also imperative to reducing the costs of waste treatment, and negative impact on the environment, and public health. Hospitals should provide their staff with sufficient knowledge and skills, as well as, facilities, to ensure that appropriate segregation and containment are implemented. More than half of the sample hospitals segregated their wastes into two types of waste: general and medical wastes. Only two hospitals mixed their general and medical wastes together, failing to segregate waste, at all. The remaining hospitals sorted their wastes into three or more categories (see Fig.1). The study indicated that waste segregation and containment in the majority of hospitals remain an issue for HCWM. They mostly segregated their wastes into two categories only, leaving the mixture of infectious and sharps wastes in the same containers, which were not always puncture proof. As for sharps wastes, only 13.92% hospitals weighed them, indicating that many hospitals mixed the sharps wastes with infectious wastes, to reduce the provision of sharps containers. From the survey, the segregation practices can be divided into two groups, good and bad practices, considering the importance of sharps containments, separated from other infectious wastes, to avoid sharps related injuries among workers at risk. Combining with the fact that a significant number of cleaning service workers was still involved in daily segregation activity at each ward, the magnitude of the problem cannot be overlooked. Thus, remedial action is necessary, and importantly, action focusing on policies, and formulating codes of conduct for each stakeholder within hospitals. This is significantly different from the results of another study [16], which found that most hospital staff (doctors and nurses) knew about the different handling 139

6 between general and medical wastes, and they segregate them appropriately in color coded bins (96.00%) as such facilities were readily available No Segregation Segregated into 2 types Segregated into 3 types Segregated into 4 types Segregated into 5 types Segregated into general & 8 others Fig. 1: Categories of solid waste segregation at source Using color-coded bags and bins as required by the Ministerial Decree No. 1204/2004 is also an important variable of compliance [17]. The study revealed that more than half the hospitals used the color-coded system. About half of them did not, due to the unaffordable costs of color-coded plastics and bins. Similarly, most of them also contained sharps wastes in plastic containers, followed by safety boxes, and plastic bags. The use of plastic bags for sharps wastes is of concern as it leads to injuries. Figure 2 shows the proportion of hospitals providing color-coded containers as required by the Health Ministerial Decree No. 1204/2004. Only 32.22% hospitals used color-coded containers and plastic bags for their wastes % 33.33% Hospital fully categorized 35.44% Hospital partly categorized Hospital does not categorized Fig. 2: Proportions of color-coded containment in sample hospitals Determinants of segregation at source and color-coded collection system The logistic regression model was used because the dependent variables intended to represent compliance are categorical in nature. Its purpose was to predict discrete outcomes - whether a hospital segregates into more than three types or less/not at all (SEGREGATION), and whether a hospital fully categorized waste by color code or partly/not at all (COLCODING) - from a set of explanatory variables. Two models were estimated and analyzed: Waste Segregation Model and Color Coding Model. 140

7 The first model attempts to determine the factors that influence the likelihood of a hospital in segregating its waste into three types or more. There are nine explanatory variables. Of those nine, four are statistically significant. They are WMBUDGET, WMPLAN, CENPOLICY, and HOSPGUIDE. A hospital that has a routine budget for HCWM is three times more likely to segregate its waste into three or more types than a hospital that does not, implying that hospitals with a routine budget can provide adequate plastic bags and bins for segregation practices. Moreover, hospitals with waste management plans are 30% more likely to segregate waste into three types or more, than hospitals without such plans. Therefore hospitals with routine budgets and waste management plans are more likely to provide staff with facilities and SOPs to segregate waste into three or more categories. Having hospital guidelines on HCWM implementation, however, has the inverse effect. A hospital issuing such guidelines is three times less likely to segregate its waste into three or more types than a hospital that does not. A possible explanation is that having guidelines is not enough; the provision of adequate facilities for segregation practices is necessary to complement such guidelines. Table 4: Estimated parameters of the waste segregation model Variables β (SE) OR (95% CI) Dependent Var. = SEGREGATION (3 or more types vs. 2 types or no segregation) 141 p-value Intercept 1.50 (0.65) DWMBUDGET (Ref: No) Yes 1.02 (0.50) 2.76 ( ) DWMUNIT (Ref: No) Yes 0.43 (0.38) 1.54 ( ) DWMPLAN (Ref: No) Yes 1.31 (0.42) 3.69 ( ) DCENPOLICY (Ref: No) Yes (0.41) 0.26 ( ) DMANPOLICY (Ref: No) Yes 0.54 (0.37) 1.71 ( ) DHOSPGUIDE (Ref: No) Yes (0.00) 0.34 ( ) DROOMSOP (Ref: No) Yes 0.71 (0.41) 2.03 ( ) DHOSPCLASS (Ref: Class C or D) Class A or B 0.47 (0.33) 1.61 ( ) DLOCATION (Ref: Outside of Java and Bali island) In Java or Bali island 0.04 (0.33) 1.04 ( ) Note: SE = standard errors; OR = odds ratio; CI = Confidence Interval; R 2 =

8 The second model seeks to determine the factors that affect the likelihood of a hospital categorizing its waste according to color codes. Of nine predictors in this model, four variables are statistically significant; they are: ROOMSOP, WMUNIT, HOSPCLASS, and LOCATION. Having the SOPs in the segregation room has considerable influence. A hospital with such SOPs is seven times more likely to fully categorize its waste by color codes than a hospital without. This indicates that in-room SOPs will guide any staff performing medical services in containing their wastes in appropriate bins or containers. Similarly, a hospital with a waste management unit is three times more likely to segregate its wastes into three or more categories, since such units are accompanied by sufficient funds and facilities that encourage segregation at source. In contrast, a hospital without a waste management unit is less likely to segregate its wastes into three or more categories, as the hospital does not have a budget for waste segregation facilities. Table 5: Estimated parameters of the color-coding model Variables β (SE) OR (95% CI) Dependent Var. = COLCODING (Fully categorized vs. Partly categorized or not at all) 142 p-value Intercept (1.08) - p<0.001 DWMBUDGET (Ref: No) Yes 0.82 (0.65) 2.28 ( ) DWMUNIT (Ref: No) Yes 1.16 (0.48) 3.19 ( ) DWMPLAN (Ref: No) Yes (0.47) 0.53 ( ) DCENPOLICY (Ref: No) Yes 0.25 (0.48) 1.28 ( ) DMANPOLICY (Ref: No) Yes 0.46 (0.42) 1.58 ( ) DHOSPGUIDE (Ref: No) Yes 0.72 (0.68) 2.05 ( ) DROOMSOP (Ref: No) Yes 1.97 (0.66) 7.16 ( ) DHOSPCLASS (Ref: Class C or D) Class A and B 0.91 (0.35) 2.49 ( ) DLOCATION (Ref: Outside of Java or Bali island) In Java or Bali island 1.16 (0.39) 3.18 ( ) Note: SE = standard errors; CI = Confidence Interval; Ref. = reference category; R 2 = The next variable, hospital location, influences segregation practices, e.g., a hospital in Java-Bali islands is three times more likely to fully categorize its waste streams, using color

9 codes provided by the relevant decree, than a hospital located outside Java-Bali islands. This points to the likelihood that a hospital in Java-Bali islands has sufficient resources for segregation facilities, than one outside Java-Bali islands. Finally, a higher class hospital is twice as more likely to fully categorize its waste using color codes, indicating better resources that enable their staff to segregate wastes into three or more categories. There are no comparable studies that have statistically analyzed as many variables. Consequently, this study adds significant findings of determinants of HCWM in developing countries like Indonesia. These findings can also be used to develop policy framework including management aspects of HCWM which in turn contributing to the reducing of HAI and waste related illnesses. CONCLUSIONS This study shows that considerable number of hospitals in Indonesia remains mix general and medicals wastes, putting hospital communities and general public at a high risk of HAI and waste related diseases. Combining with inadequate of collection facilities including color-coded plastic lids and containers, the implementation of safe HCWM was constrained. Since the practices of waste segregation at the point of generation is of importance as it is the core phase of good WMH, the above study findings are very important as to develop sustainable HCWM. Factors which are determined in the statistical models can be utilized by hospital managers and other stakeholders to establish a comprehensive hospital waste management planning to improve the current status of HCWM which is in fact far from satisfactory. These include the availability of waste management policy either at the central level or at the provincial/district levels, availability of waste management plans, guidelines, SOPs, and sufficient budgets. Acknowledgements: The authors acknowledge the provision of research fund from National Institute of Health Research and Development, Indonesian Ministry of Health and the assistance of hospital managers of all sample hospitals who were willing to participate in this research. References 1. MoH, National health profile Ministry of Health (MoH), Jakarta. 2. Prüss, A., Giroult, E., & Rushbrook, P. (Eds.), Safe management of wastes from health-care activities. WHO. Geneva. 3. Sasimartoyo, T. P Report on the assessment of hospital sanitation in Indonesia, Jakarta: Ministry of Health and WHO. Jakarta. 4. Bi, P, P. J. Tully, S. Pearce, &, J. E. Hiller, Occupational blood and body fluid exposure in an Australian teaching hospital. Epidemiology and Infection, 134(3): Tudor, T. L., S. W. Barr, & A. W. Gilg, Linking intended behaviour and actions: A case study of healthcare waste management in the Cornwall NHS. Resources Conservation and Recycling, 51(1): Cheng, Y. W., F. C. Sung, Y. Yang, Y. H. Lo,, Y. T. Chung, & K. C. Li, Medical waste production at hospitals and associated factors. Waste Management, 29(1): Irianti, S., & S. Herat, Sustainable Health Care Waste Management in Indonesia: Existing Problems and Proposed Solutions. International Conference on Sustainable Environmental Technology and Sanitation for Tropical Region, Surabaya, 19 November, MoH, Indonesian Health Profile Ministry of Health (MoH), Jakarta. 9. Diaz, L. F.,, L. L. Eggerth, S. Enkhtsetseg, &, G. M. Savage, Characteristics of healthcare wastes. Waste Management, 28(7):

10 10. Mbongwe, B., B. T. Mmereki, & A. Magashula, Healthcare waste management: Current practices in selected healthcare facilities, Botswana. Waste Management, 28: Tudor, T. L.,, C. L. Marsh, S. Butler, J. A. Van Horn, & L. E. T. Jenkin, Realising resource efficiency in the management of healthcare waste from the Cornwall National Health Service (NHS) in the UK. Waste Management, 28(7): WHO Health-care waste management: rapid assessment tool. World Health Organization (WHO), Geneva. 13. Hill, R. C., W. E.Griffiths, & G. C. Lim Principles of econometrics. Hoboken, NJ: John Wiley & Sons, Inc. 14. Da Silva, C. E., A. E. Hoppe, M. M. Ravanello, & N. Mello, Medical wastes management in the south of Brazil. Waste Management, 25(6): Taghipour, H., & M. Mosaferi, Characterization of medical waste from hospitals in Tabriz, Iran. Science of the Total Environment, 407(5): Ramokate, T., & D. Basu, Health care waste management at an academic hospital: Knowledge and practices of doctors and nurses. South African Medical Journal, 99(6): MoH Health Ministerial Decree on Hospital Environmental Health Standars. Ministry of Health (MoH), Jakarta. 144