IHRC is a for profit staffing company. Government funds are put into an IHRC contract so that IHRC will supply trained individuals to accomplish the projects written into the contract. The individuals then work on site at CDC with other CDC staff members to accomplish project goals. The IHRC status does not alter the original agreement signed for PLOS nor does it set up competing interests. This does not alter the authors‚ adherence to all the PLOS ONE policies on sharing data and materials.
Conceived and designed the experiments: EA PB SC WB AC MB JB. Performed the experiments: EA PB SC WB AT MC MB JB. Analyzed the data: EA PB MS SC AT. Wrote the paper: EA PB MS SC MB JB.
In 2008, a large
Community-wide outbreaks associated with public drinking water systems are rare in the United States since drinking water regulations were implemented by the Environmental Protection Agency (EPA), beginning in 1974 with the Safe Drinking Water Act (SDWA)
Alamosa is a small municipality of approximately 8,800 residents situated between two mountain ranges in the San Luis Valley of south-central Colorado
As a result of the outbreak, 434 cases, including 124 laboratory-confirmed cases, 20 hospitalizations, and one death were reported; a telephone survey conducted by the Colorado Department of Public Health and Environment (CDPHE) at the time of the outbreak indicated that an estimated 1,300 persons became ill (CDPHE, unpublished data). Anecdotal reports of subsequent complications due to
An extensive investigation conducted at the time of the outbreak involved multiple agencies. Water supply interruptions necessitated a large-scale response from local, state, and federal agencies, including the Colorado National Guard, and volunteer agencies. The economic burden to the community was thought to be significant due to business and school closures, missed work to care for ill family members, and the costs of obtaining potable water and other supplies. Because of the scope and extent of the outbreak and response, CDPHE and the local health department in Alamosa requested assistance from the Centers for Disease Control and Prevention (CDC) to assess the full economic and long-term health impacts on the community of Alamosa.
This data collection was judged by officials at CDC to be non-research public health practice, and therefore was not subject to Institutional Review Board (IRB) review. Nevertheless, written informed consent was obtained from all participants, and participants were given the option to refuse specific questions or to decline responding to the surveys.
We conducted a community-wide household survey to assess the health and economic impacts of the outbreak; a business survey to assess the costs incurred by businesses; school surveys to document closures and costs; a review of billing data from two local health care systems to assess direct healthcare expenditures; and interviews with local and state governmental and non-governmental agencies to document costs related to the emergency response efforts.
In October 2009, we sent or hand-delivered a survey to all households that received a water bill from the City of Alamosa (as of September 2009) and surveys were returned via postal mail to CDC. Survey questions covered topics such as residents’ drinking water source before and after the outbreak, alternate water sources used during the outbreak, household illness during the outbreak, including potential long-term health consequences (e.g., joint, skin, urinary tract, eye, or other problems occurring within one month following diarrheal onset), and other demographic and household characteristics. Households were also asked to report economic costs associated with the outbreak, including costs associated with illness (e.g., over the counter medicine and out-of pocket costs for prescription medications, doctor’s visits and hospitalization), caring for ill family members, securing alternate water sources (bottled water or water filters). To calculate indirect cost of illness, ill household members and caretakers were also asked to provide information on their occupation and daily wage (see Table S1 in File S1 for more information). Some questions were posed at the household-level (e.g., costs for purchase of bottled water) while others were reported for each member of the household (e.g. symptoms, occupation, and demographic characteristics). For our analysis, in order to be consistent with the case definition used during the outbreak, a case was defined as anyone who reported diarrhea (≥2 loose stools during a 24 hour period) during the outbreak. An affected household was any household with ≥1 person who experienced diarrhea during the outbreak.
To describe how the outbreak impacted local businesses financially, in October 2009 we sent a survey to all businesses inspected by CDPHE (N = 128), including retail food establishments (restaurants, hotels, nursing homes, and child care centers), and other businesses (N = 54) potentially affected by the water shortages (grocery stores, beauty salons, dentists, and animal clinics). Contact information was provided by CDPHE, or obtained through a telephone directory and internet searches. Businesses were asked how the outbreak affected their business, including whether the business had to close, lay off workers, and whether the business had to buy additional water or ice, lost or gained money overall, and if the business ever regained pre-outbreak levels. To encourage businesses to respond and protect confidentiality, we did not ask for the business name or address on the survey. Five of the 128 CDPHE-inspected businesses were located outside of Alamosa but had clientele likely to be comprised of Alamosa residents.
We interviewed via telephone or in-person staff from the City of Alamosa, Alamosa County Nursing Service, and CDPHE and the local chapter of the American Red Cross to ascertain estimates of the direct and indirect cost of the outbreak response to local and state governmental and non-governmental agencies. Respondents provided information on the cost of the response (e.g., lodging and meals for staff, truck rentals, etc.), the number of staff and their aggregate labor hours spent responding to the outbreak. We also interviewed Alamosa health care providers, including a hospital, medical practices, nursing homes, and assisted living facilities to assess the outbreak impact and to request billing data to supplement cost estimates from the household surveys (see Supporting Information and Table S2 in File S1for more information). To determine the effects of the outbreak on educational institutions, we interviewed representatives from each of Alamosa’s two public colleges, two private schools, and its public school district. School representatives were asked about the types of additional costs incurred because of the outbreak, including purchasing bottled water or hand sanitizer, paying for employee overtime, and costs for make-up days.
All survey data were entered into a Microsoft Access 2007 database and descriptive analyses were conducted using SAS v. 9.2 (Cary, NC). We compared survey respondents’ characteristics (sex, age, race/ethnicity and income) to the characteristics of the 2008 City of Alamosa population
For our cost estimates, we took a societal perspective and defined costs as expenses which would not be incurred if the outbreak had not occurred. Since almost all costs were incurred in 2008 and 2009, we did not apply a discount rate. No capital costs (materials with more than a 5 year useful life) were incurred. All the costs were recorded in 2008 U.S. dollars.
We built a Monte Carlo simulation model using @Risk software (Palisade Corporation, NY) to extrapolate the costs to the city of Alamosa. The model used the following formula (see Table S3 and Figure S1 in File S1 for more information):
Cost distribution of the given cost (from household survey).
We assumed that the proportion of respondents from the survey who experienced a given cost was the same proportion in the community who would have experienced the costs. We used the costs reported by all individuals/households in the survey to generate the cost distribution to fit the data for the model. The details of these cost distributions are given in the Supporting Information (Table S3 in File S1). Using Monte Carlo simulation (10,000 iterations), we then extrapolated the costs to the city of Alamosa and the model’s results are presented as the total cost (the median of the 10,000 iterations) and range (representing the 5th to 95th percentile of the 10,000 iterations). Additional methodological details about the model, methods for direct and indirect cost calculations, and extrapolation methods, are provided in the Supporting Information (see File S1).
The community survey was distributed to all households that received municipal drinking water (N = 2,692). After excluding non-responders and ineligible responses (refusal, out of town during the outbreak, not on city water, or other reasons) 29% (n = 771) of households, representing 1,732 persons, returned surveys eligible for analysis (
Approximately one-third (242/771, 31%) of households, and 21% (369/1732) of individual respondents, reported diarrheal illness during the outbreak. Fifty-seven percent (n = 187/329) of ill persons were female (including four who were pregnant); the median age of ill persons was 37 years (range: 0–98 years). By definition, all ill persons experienced diarrhea out of which 30% (n = 110) reported bloody diarrhea. The median duration of illness was four days (range: 1–60 days). In total, survey respondents (n = 350) were sick for 2,341 person-days. Most (n = 194, 80%) of the Alamosa households with at least one ill person reported buying medicine or other items because of their illness (
Household Survey | City of Alamosa (2008) | ||||||
(n = 1,732 individuals, 771 households) | (N = 8,746 persons, 3,302 households) | ||||||
Cost($) | |||||||
Reported in survey |
Simulation model |
||||||
n(%)incurred cost | n(%)reported cost |
Mean | Median | Total | Total | (range) | |
|
|||||||
Bought bottled water | 657(85%) | 523(68%) | $87 | $50 | $45,530 | $135,781 | ($30,498–$604,183) |
Bought water filtration system | 202(26%) | ||||||
Installation of filter | 202(26%) | 195(25%) | $180 | $70 | $35,084 | $83,536 | ($7,760–$336,936) |
Maintenance of filter | 202(26%) | 168(22%) | $121 | $60 | $20,343 | $53,715 | ($3,974–$232,136) |
Stay overnight somewhere else | 123(16%) | 90(12%) | $362 | $233 | $32,549 | $113,266 | ($31,749–$403,950) |
|
|
|
|
||||
|
|||||||
Bought items or received medical treatment | 194(80%) | (0%) | |||||
Bought nonprescription medicine | 168(69%) | 162(67%) | $40 | $25 | $6,483 | $18,220 | ($4,617–$71,847) |
Bought other things (e.g., Gatorade or diapers) | 120(50%) | 116(48%) | $44 | $25 | $5,102 | $13,765 | ($4,127–$45,893) |
Went to doctor or a clinic | 75(31%) | 64(26%) | $80 | $43 | $5,116 | $12,653 | ($936–$54,660) |
Received a prescription | 32(43%) | 28(44%) | $17 | $13 | $463 | $1,850 | ($397–$5,193) |
Had diagnostic tests (e.g., blood or stool test) | 37(49%) | 35(55%) | $96 | $30 | $3,355 | $5,508 | ($408–$23,801) |
Went to hospital/emergency room | 31(13%) | 21(9%) | $320 | $40 | $6,725 | $10,603 | ($784–$45,825) |
|
|
|
|
||||
|
|||||||
|
|||||||
Work full-time | 156(42%) | 87(24%) | $430 | $300 | $2,917 | $183,644 | ($45,663–$480,473) |
Work part-time | 102(28%) | 8(2%) | $215 | $150 | $37,889 | $47,083 | ($11,716–$123,194) |
Non-worker | 111(30%) | (0%) | $0 | $0 | $0 | $0 | (n/a) |
|
|||||||
Paid caretakers | 11(10%) | 7(7%) | $178 | $110 | $1,245 | $4,676 | (n/a) |
Unpaid caretaker (work full-time) | 77(73%) | 59(56%) | $913 | $588 | $54,165 | $148,173 | ($31,958–$686,232) |
Unpaid caretaker (work part-time) | 15(14%) | 1(1%) | $457 | $294 | $640 | $14,434 | ($3,115–$66,844) |
Unpaid caretaker (non-worker) | 3(3%) | (0%) | $0 | $0 | $0 | $0 | (n/a) |
|
|
|
|
||||
|
|
|
|
Respondents who incurred cost
Costs extrapolated to the City of Alamosa as: total individuals/households in City of Alamosa (from census)×% incurring costs (column 3 above)×cost distribution (Table S3 in File S1) using a Monte Carlo simulation model with 10,000 iterations. Total cost derived from median of 10,000 iterations and range represents the 5th to 95th percentiles of the 10,000 iterations of the Monte Carlo simulation model. See main text and Supporting Information (in File S1) for details.
Twenty-nine percent (n = 108) of all ill persons reported experiencing ≥1 potential long-term health consequence of
Potential long-term health consequence | n (%) | Days after diarrhea began that problem started: mean (range) |
Duration (weeks) | |
Mean (range) time-limited duration of symptoms |
Symptoms still present at time of survey n(%) | |||
Rash, itchiness or other skin problems | 52 (14%) | 5 days (0–30) | 3 weeks (1–24) | 11/52 (21%) |
Arthritis, aching joints or other joint problems | 51 (14%) | 7 days (1–30) | 3 weeks (0–16) | 20/51 (39%) |
Urinary tract problems (e.g., pain or burning during urinationor a discharge) | 32 (9%) | 7 days (1–30) | 4 weeks (1–30) | 5/32 (16%) |
Eye problems such as pain or redness | 19 (5%) | 4 days (1–7) | 2.5 weeks (1–6) | 2/19 (11%) |
Abscess (skin, soft tissue, anal, etc.) | 6 (2%) | 5 days (2–14) | 2 weeks (1–3) | 1/6 (17%) |
Other serious complications (e.g., bowel perforation or peritonitis, septic arthritis, or endocarditis) | 7 (2%) | n/a |
n/a |
n/a |
As reported by 43/52 with skin problems, 36/51 with joint problems, 27/32 with urinary tract problems, 16/19 with eye problems, and 5/6 with abscesses.
As reported 36/52 with skin problems, by 22/51 with joint problems, 15/32 with urinary tract problems, 13/19 with eye problems, and 4/6 with abscesses.
Questions were not asked.
During the bottled water advisory, Do Not Use order, and boil water advisory, most households reported using bottled water (either purchased or donated) for drinking, cooking and brushing teeth (
Avoided activity | Bought bottled water | Used bottled/bulk water given out for free | Boiled tap water | Used water from outside Alamosa |
Used treatedtap water |
Used un-boiledtap water | Used water from other source | |
|
9(1%) | 602(78%) | 556(78%) | 90(12%) | 228(30%) | 21(3%) | 14(2%) | 5(1%) |
|
33(4%) | 465(60%) | 531(60%) | 233(30%) | 238(31%) | 28(4%) | 42(5%) | 9(1%) |
|
121(16%) | 272(35%) | 368(35%) | 330(43%) | 225(29%) | 59(8%) | 158(20%) | 35(5%) |
|
5(1%) | 518(67%) | 497(67%) | 107(14%) | 194(25%) | 20(3%) | 54(7%) | 6(1%) |
|
121(16%) | 107(14%) | 189(14%) | 134(17%) | 427(55%) | 27(4%) | 366(47%) | 23(3%) |
Questions allowed for multiple options and therefore row totals do not sum to 100%.
e.g., a friend’s house, hotel, or artesian spring.
e.g., using chlorine or a filter.
Bottled Water Advisory |
Do Not Use Order |
Boil Water Advisory |
|
(March 19–24, 2008) | (March 25–April 3, 2008) | (April 3–11, 2008) | |
|
71(9%) | 25(3%) | 58(8%) |
|
212(28%) | 73(9%) | 233(30%) |
|
414(54%) | 191(25%) | 456(59%) |
|
145(19%) | 58(8%) | 116(15%) |
|
516(67%) | 256(33%) | 498(65%) |
During the bottled water advisory, residents were told to use bottled water for drinking, cooking, brushing teeth, and dishwashing but that if no bottled water was available they could boil their water.
During the Do Not Use order, while the distribution system was being hyperchlorinated, residents were told to only use their tap water for flushing toilets.
During the boil water advisory, residents were told to boil their water before using it for drinking, cooking, or brushing teeth.
Most households (n = 699, 91%) reported municipal tap water as their main drinking water source at home prior to the outbreak (
Switched from tap to bottled water (n = 249) |
Added a new filter (n = 114) |
|
|
134(55%) | 75(66%) |
|
110(45%) | 42(37%) |
|
60(25%) | 35(31%) |
|
29(12%) | 16(14%) |
Categories are not mutually exclusive so percentages can sum to >100%.
Outbreak-related costs for City of Alamosa residents totaled an estimated $846,907 (range: $182,468–$3,185,843) (
We distributed 177 surveys to establishments inside the City of Alamosa and 5 to businesses located outside the City of Alamosa that primarily served or employed Alamosa residents. Of the 182 surveys, 21 (12%) were undeliverable and 50 (50/161, 31%) were returned. The following surveys were excluded: three because the business was not open during the outbreak and one because it was a city jail. Forty-six (46/161, 29%) surveys were eligible for analysis: 41 surveys were included in the primary analysis and an additional five surveys were analyzed separately because the business was either located outside Alamosa or not connected to municipal water.
The 41 businesses located in Alamosa and connected to municipal water included retail stores (n = 8, 20%), restaurants or other food service establishments (n = 6, 15%), beauty salons or barber shops (n = 6, 15%), child care centers (n = 4, 10%), nursing homes or long-term care facilities (n = 3, 7%), and other types of businesses (n = 10, 24%); two (5%) businesses did not specify the type of establishment. One-third (14/41, 34%) of responding businesses closed during the outbreak (mean length of business closure: 8.4 days). Approximately half of businesses reported losing money due to the outbreak, with a median loss of $8,750 (range: $400–$200,000) (
Business Survey (n = 46) | Extrapolated to Sample ofCity of Alamosa Businesses (N = 156) | ||||
Reported in survey | Estimated using simulation model |
||||
n(%) | Mean | Median | Total (range) | ||
|
|||||
Lost money | 22(54%) | $35,306 | $8,750 |
|
|
No change | 17(41%) | ||||
Did better because of the outbreak | 2(5%) | ||||
|
|||||
Lost money | 1(20%) | $13,967 | $13,967 |
|
|
No change | 3(60%) | ||||
Did better because of the outbreak | 1(20%) | ||||
|
|
Costs extrapolated to the City of Alamosa as: total businesses on municipal water (N = 156)×% incurring costs (column 3 above)×cost distribution (Table S3 in File S1) using a Monte Carlo simulation model with 10,000 iterations. Total cost derived from median of 10,000 iterations and range represents the 5th to 95th percentiles of the 10,000 iterations of the Monte Carlo simulation model. See main text and Supporting Information (in File S1) for details.
Outbreak response cost estimates to local, regional, and state governmental and volunteer organizations totaled $823,314 (
Personnel | Other |
Total | |
|
|||
Federal Government | $19,040 |
$19,040 | |
State of Colorado | $215,925 | $316,449 |
$532,374 |
Alamosa County | $52,817 | $7,582 | $60,399 |
City of Alamosa | $50,872 | $19,023 | $69,895 |
|
|||
Volunteer organizations | $40,135 |
$40,575 | $80,710 |
Other organizations | $60,896 |
$60,896 | |
|
|
|
|
Includes transportation, supplies, lodging, etc.
As captured by City of Alamosa record-keeping.
Includes expenses covered by the state disaster fund (e.g., National Guard, incident management teams); these were all included in the “Other” category because personnel costs were not reported separately from other expenses.
Includes $7,920 for staff overtime (3 persons and a total of 750 hours of overtime) and $32,215 in estimated indirect costs associated with volunteer time (4,589 volunteer hours estimated at Colorado minimum wage rate of $7.02/hour).
Most health care providers reported significant expenses in securing and providing clean water (for drinking, bathing, housekeeping and other uses) or disposable supplies. However, these costs could not be estimated because most could not retrospectively itemize these expenses. Only one local hospital was able to provide billing records for outbreak-related care it provided to 104 of the 124 laboratory-confirmed cases. The estimated total cost of health insurance payments for Alamosa City residents that sought health care was $244,985 (range: $65,615–$928,915) (
Household Survey | Hospital A Cost Estimates | Extrapolated to City of Alamosa | ||||||||
(n = 1,732) | (n = 104 culture-confirmed casesand 139 separate healthcare visits) | (N = 8,746) | ||||||||
Cost ($) |
||||||||||
n(%) | n(%) | Mean | Median | N(%) |
Total (range) | |||||
Ill Persons | 369(21%) | 1,423(16%) | ||||||||
Sought care | 107(29%) | 413(29%) | ||||||||
Clinic/doctors’ office |
76(71%) | 67(48%) | $129 | $93 | 293(71%) |
|
||||
Emergency department | 22(21%) | 67(48%) | $693 | $390 | 85(21%) |
|
||||
Hospitalized | 9(8%) | 5(4%) | $7,011 | $3,159 | 35(8%) |
|
||||
|
|
|
Because data were obtained from the hospital only, clinic/doctor’s office visit costs only include laboratory but not physicians’ fees.
We have removed the background rate of diarrhea in the population (5%) to get the percent of illness due to outbreak (21%−5% = 16%). The number of ill persons was the denominator for subsequent proportions who incurred the costs (e.g., 29%,71%, 21% and 8%).
Costs extrapolated to the City of Alamosa as: total population (N = 8,746)×% incurring costs (column 3 above)×cost distribution (Table S3 in File S1) using a Monte Carlo simulation model with 10,000 iterations. Total cost derived from median of 10,000 iterations and range represents the 5th to 95th percentiles of the 10,000 iterations of the Monte Carlo simulation model. See main text and Supporting Information (in File S1) for details.
Total | (range) | % of total | |
|
|
|
|
Outbreak-related expenses | $386,298 | ($73,981–$1,577,205) | |
Direct out-of-pocket health care costs | $62,599 | ($11,269–$247,219) | |
Indirect costs of acute illness and caretaking | $398,010 | ($97,218–$1,361,419) | |
|
|
|
|
|
|
|
|
|
|
|
|
Federal Government | $19,040 | n/a | |
State of Colorado | $532,374 | n/a | |
Alamosa County | $60,399 | n/a | |
City of Alamosa | $69,895 | n/a | |
|
|
|
|
Volunteer organizations | $80,710 | n/a | |
Other | $60,896 | n/a | |
|
|
|
|
|
|
|
|
|
|
|
|
Details provided in
The total estimated economic impact of the outbreak, including costs to City of Alamosa residents, businesses, schools, and healthcare facilities and the governmental and non-governmental outbreak response was approximately $2.6 million (range: $1.1 million–$7.8 million dollars) (
Since passage of the Safe Drinking Water Act and its amendments by EPA
Despite our comprehensive approach, this outbreak cost estimate is lower than previous epidemiologic studies of outbreaks in public water systems, perhaps due to our conservative methodological approach and the differences in the size of the affected population or duration of the outbreak. Harrington et al. estimated the economic impact of a 1984 waterborne outbreak of giardiasis in a Pennsylvania county at $18.2–133.3 million (in 2008 dollars) and did not include the cost of the outbreak response or the impact on local businesses
In our assessment, 31% of households and 21% of survey respondents became ill during the outbreak. Approximately one-third of those who became sick reported a potential long-term health consequence following their diarrheal illness and, of those, 26% were still experiencing symptoms 18 months after the outbreak. Because all symptoms were based on self-report, and may have been coincidental to, rather than caused by the
Over 90% of households reported that municipal water was their main drinking water source at home prior to the outbreak. After the outbreak, 38% of respondents mainly drank bottled water and only 30% of households continued to primarily drink tap water; an additional 15% purchased a new filter or filtration system. The purchase of bottled water and installation and maintenance of filters cost City of Alamosa residents approximately $273,000 during the outbreak. Almost half (45%) of survey respondents cited safety concerns as a reason for switching from tap to bottled water. This lack of trust was also apparent in survey participants’ comments, such as: “I will never again fully trust the system or drink any tap water without some concern…” and “I still don’t feel safe drinking or cooking with the city water… I have spent a lot of money buying bottled water.”
The economic impact of the outbreak on the sample of businesses was one of the largest expenses, totaling $626,000 and accounting for 24% of the total outbreak costs. Approximately half of businesses that responded indicated that they lost money and approximately one-third had to close temporarily during the outbreak. Only 60% reported ever returning to pre-outbreak financial levels, including one that noted that “it took 2–3 months to get back to previous levels.” Because the survey was sent 18 months after the outbreak, it could have failed to reach businesses that might have been forced to close because of the outbreak. Household survey responses corroborated this; one respondent noted that “we couldn’t pay our mortgage [and] lost our restaurant. We now both work for someone else for not as much pay. We had our restaurant for 25 years.”
This assessment was subject to several limitations. First, our outbreak cost estimate is likely an underestimate. It does not include health care costs for individuals who sought care outside of Alamosa (either because some ill individuals may not have responded to the survey or because we were only able to obtain hospital-associated costs from one local hospital). We also were unable to assign an estimate for the one death associated with the outbreak. Alamosa is the geographic and commercial center of the San Luis Valley, and many people from surrounding areas work and dine in Alamosa but the survey did not capture business-related costs or health impact for people who live outside of Alamosa or for businesses that either did not receive a survey or did not respond. Additionally, outbreak response costs incurred by the federal government and by local organizations or municipalities outside the City of Alamosa that contributed to the outbreak response are likely incomplete. Household survey respondents also mentioned various costs not covered in the questionnaire, such as the cost of gas, disposable plates/utensils, or pet care associated with the outbreak. Additionally, direct and indirect costs associated with the long-term health consequences of
Second, we assumed that the survey respondents were a representative sample of the City of Alamosa population, yet our survey respondents differed by age, sex, ethnicity, and socioeconomic status
The likely source of the outbreak was determined to be animal contamination of a storage tank that had numerous cracks and entry points
(DOCX)
The authors gratefully acknowledge the contributions of all of those who aided in the outbreak response and in this investigation, particularly the Alamosa County Nursing Service and the American Red Cross.
The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the Centers for Disease Control and Prevention.