The authors have declared that no competing interests exist.
Conceived and designed the experiments: RS SPL NN ESG. Performed the experiments: NAR RS MSI MU. Analyzed the data: NAR RS. Wrote the paper: NAR RS SPL MSI MU MJH RUZ NN ESG.
This paper describes the physical structure and environmental contamination in selected hospital wards in three government hospitals in Bangladesh.
The qualitative research team conducted 48 hours of observation in six wards from three Bangladeshi tertiary hospitals in 2007. They recorded environmental contamination with body secretions and excretions and medical waste and observed ward occupant handwashing and use of personal protective equipment. They recorded number of persons, number of open doors and windows, and use of fans. They measured the ward area and informally observed waste disposal outside the wards. They conducted nine focus group discussions with doctors, nurses and support staff.
A median of 3.7 persons were present per 10 m2 of floor space in the wards. A median of 4.9 uncovered coughs or sneezes were recorded per 10 m2 per hour per ward. Floors in the wards were soiled with saliva, spit, mucous, vomitus, feces and blood 125 times in 48 hours. Only two of the 12 patient handwashing stations had running water and none had soap. No disinfection was observed before or after using medical instruments. Used medical supplies were often discarded in open containers under the beds. Handwashing with soap was observed in only 32 of 3,373 handwashing opportunities noted during 48 hours. Mosquitoes and feral cats were commonly observed in the wards.
The physical structure and environment of our study hospitals are conducive to the spread of infection to people in the wards. Low-cost interventions on hand hygiene and cleaning procedures for rooms and medical equipment should be developed and evaluated for their practicality and effectiveness.
Hospital-acquired infection represents a major public health concern worldwide. Hospitals have played a significant role in the spread of emerging infections. In Toronto, 77% of case patients were exposed to severe acute respiratory syndrome (SARS) in hospital settings in 2003
Pathogens in hospital environments can be transmitted through airborne particles, fomites, respiratory droplets or direct contact with bodily fluids
The study protocol was approved by the Ethical Review Committee of the International Centre for Diarrhoeal Disease Research, Bangladesh (FWA # 00001468, Human Welfare Assurance # 00001822). The team obtained informed consent from hospital authorities for data collection and secured written consent from participants before conducting discussions. They observed public behavior and individuals were not identified.
The main methods of data collection for this exploratory qualitative study were structured and semi-structured direct observation. A team of three anthropologists and two sociologists, trained in qualitative research methods, collected data from March through September 2007 from one pediatric and one adult male medicine ward from each of three public tertiary teaching hospitals. They mapped the wards to describe physical layout and calculated the floor area. Next, they conducted 48 hours of observation in 22 sessions: three to four sessions in each ward. To capture variation in activities at different times of day, sessions were held during three non-overlapping periods; three hours from 9∶00 am–2∶30 pm, three hours from 3∶30 pm–9∶00 pm and one hour from 10∶00 pm–12∶30 am. Through structured direct observation, they recorded number of ward occupants-including patients, family caregivers, visitors and healthcare workers, use of fans, and number of open doors and windows at the beginning and end of each session. They also recorded frequencies of coughing and sneezing on the wards, use of personal protective equipment and handwashing. They recorded handwashing opportunities, defined as events during which ward occupant hands were contaminated with body secretions or excretions. Handwashing opportunities included points at which hands should have been washed before an activity, such as providing patient care
To complement observation findings, the team conducted nine focus group discussions; one with each of the groups of doctors, nurses and support staff in each hospital. They approached all the staff working in the study wards and some staff from other wards and enrolled those who consented to participate in the discussions. Each discussion included six to 11 participants and lasted for 45 to 80 minutes. The discussions were facilitated at the hospitals by NAR, RS and MSI and audio recorded.
The team expanded the observation field notes and transcribed the recorded data verbatim from discussions. NAR and RS reviewed data from observations to identify emerging themes relevant to the study objective and summarized the data according to those themes. They also reviewed focus group discussions to identify relevant data to further cross-check and complement the observations.
The wards had either an open floor plan or cubicles with four-foot high walls. The floor areas ranged from 101 to 317 m2 (
Characteristics | Hospital A | Hospital B | Hospital C | |||
Adult | Pediatric | Adult | Pediatric | Adult | Pediatric | |
Ward area for patients (m2) |
317 | 241 | 260 | 101 | 205 | 175 |
Number of beds | 35 | 31 | 30 | 15 | 30 | 33 |
Mean distance between beds (m) | 0.8 | 0.7 | 1.1 | 0.6 | 0.8 | 0.6 |
For doctors and nurses | 1 | 0 | 1 | 2 | 2 | 8 |
For patients, visitors and support staff | 4 | 1 | 4 | 2 | 4 | 2 |
For doctors and nurses | 1 | 0 | 1 | 2 | 2 | 8 |
For patients, visitors and support staff | 3 | 1 | 4 | 2 | 4 | 2 |
For patients, visitors and support staff | 3 | 1 | 3 | 2 | 2 | 2 |
For patients, visitors and support staff | 2 | 0 | 3 | 1 | 1 | 2 |
For patients, visitors and support staff | 2 | 2 | 2 | 0 | 2 | 0 |
For patients, visitors and support staff | 2 | 2 | 2 | 0 | 2 | 0 |
For doctors and nurses | 1 | 0 | 1 | 2 | 4 | 6 |
For patients, visitors and support staff | 2 | 2 | 2 | 2 | 2 | 0 |
For doctors and nurses | 1 | 0 | 1 | 2 | 4 | 6 |
For patients, visitors and support staff | 0 | 0 | 0 | 0 | 2 | 0 |
Doctors and nurses | 5 | – | 5 | 2 | 3 | 1 |
Patients, visitors and support staff | 17 | 48 | 25 | 27 | 14 | 63 |
Doctors and nurses | 5 | – | 5 | 2 | 1 | 1 |
Patients, visitors and support staff | – | – | – | – | 28 | – |
*Ward area included patient beds and nursing stations.
The condition of and accessibility to sanitation facilities varied for different categories of ward occupants (
Patients, family caregivers, visitors and support staff all used patient toilets and handwashing stations. Only two of 12 patient handwashing stations had running water and the team did not observe any water stored by the handwashing stations. There was no soap at any of these stations, though some patients and caregivers brought their own soap. Patients and family caregivers used bathroom facilities for bathing, and washing clothes and utensils. They used urinals only for urinating. Non-functioning toilets, bathrooms, urinals and handwashing stations were used as waste containers.
The hospital wards were crowded with patients, caregivers, visitors and staff, especially from morning to afternoon, when doctors made their rounds and patients were admitted and discharged. An overall median of 3.7 (interquartile range [IQR]: 2.0–5.3) ward occupants were present per 10 m2 of floor space (
Time slot ofobservations | Median number of people (IQR |
|||
Hospital A | Hospital B | Hospital C | Overall | |
9∶00 am to 2∶30 pm | 2.1 (1.6–2.5) | 5.0 (4.4–7.9) | 5.8 (4.6–8.9) | 4.6 (2.6–8.2) |
3∶30 pm to 9∶00 pm | 1.7 (1.2–2.1) | 3.6 (2.8–3.9) | 4.2 (1.8–6.4) | 2.4 (1.8–3.9) |
10∶00 pm to 12∶30 am | 1.9 (1.3–2.4) | 4.1 (3.7–4.4) | 3.7 (1.6–5.8) | 3.1 (1.6–4.4) |
Irrespective of time slot | 1.9 (1.4–2.4) | 4.1 (3.7–4.5) | 5.2 (2.2, 6.8) | 3.7 (2.0–5.3) |
*IQR indicates interquartile range.
Since the hospitals were also teaching facilities, medical students accompanied senior physicians on their rounds. On one ward, 37 students accompanied two doctors for almost an hour. Students stood nearby or sat on patient beds. Below is the description of a pediatric ward from the observation notes.
The ward became crowded, even the verandas were filled with patients and caregivers. Some of the patients’ mothers had another healthy child staying in the ward. On the examination bed and nursing table, doctors and nurses quickly examined two or three patients at a time, gave medicines, injections and/or nebulizers, placed and removed canula and drew blood.
Overall, family caregivers were most the numerous on our study wards in each hospital (
Open bowls or buckets under patient beds for disposing waste (used medical supplies, patient body fluids, discarded food) were emptied into larger drums once daily by the hospital cleaners. A cleaner stated,
“We have to carry the waste [in a bucket] from the third floor on our shoulder to dump on the ground (i.e., ground floor deposited directly on the ground). There is nothing [like a trolley] to help carry the bucket… Sometimes there are holes in the bucket and waste drops down on our bodies.”
The team observed children crawling on the floor and playing with used syringes with needles. Only Hospital C had separate cardboard boxes beside the nursing stations to discard used sharps; it also had a functioning incinerator. Cleaners discarded waste on open grounds adjacent to the hospital building. City corporation vehicles removed waste from these grounds once a week. The team observed young boys and women collecting used syringes and saline bags from hospital grounds. They reported they planned to resell them.
Feral cats were commonly observed in all wards, scavenging for food in cabinets and waste bins, climbing on patient beds and sleeping on patient bedding. The team also observed mosquitoes in the wards. All patients used mosquito nets at night in Hospitals A and C, but not in Hospital B.
Many windows and doors remained fully or partially closed. Some ceiling fans remained off every day due to electrical outages. Below is a description of ventilation in a pediatric ward.
All fans remained off, except the one in the nursing station, and most of the windows remained closed most of the time. Family caregivers explained that the children had colds, fever, breathing difficulty and pneumonia and that airflow was harmful to patients with such illnesses.
The team observed a total of 6,033 coughs and 75 sneezes in 48 hours of observation; a median of 4.9 (IQR: 4.1–7.4) uncovered coughs or sneezes per 10 m2 per hour per ward. Only 60 coughs and 20 sneezes were covered; four coughs were covered by a cloth and the rest by the cougher’s or sneezer’s hands. No persons were observed to wash hands after coughing or sneezing. Only one family caregiver used a cloth mask while caring for the patient.
The floors, walls, grills of windows and verandas, bedrails, nursing tabletops, bedcovers, mattresses and blankets were soiled with ward occupant body secretions and excretions. The team observed floors being soiled 125 times in 48 hours. The following excerpts illustrate the soiling of surfaces.
While setting a blood transfusion bag in a patient’s hand, the doctor accidentally dripped blood on the bed and floor. No staff cleaned it during observation hours.
While drawing gastric fluid from a patient’s stomach, fluid dripped on the bedcover and the stain remained visible two days later when the bed with the unchanged bedcover was occupied by another patient.
Cleaners swept the ward floors daily with dry brooms. Although cleaners in all hospitals reported wet mopping wards two to three times daily, the team observed daily wet mopping only in Hospital A and weekly sweeping with water in Hospital B. Cleaners also reported using disinfectant while mopping if disinfectants were available. Soiled blankets or mattresses were only shaken to remove dust after being used by one patient and then provided to another. No cleaning of window grills, bedrails, cabinets or walls was observed.
The floors of the patient toilet areas were wet, slippery and soiled with body secretions and excretions and food remnants. In one hour, the team observed seven pediatric patients urinate on the floor near the entrance of the toilet area at night since there was no light inside the toilet area.
Doctors and nurses used the same medical instruments, such as stethoscopes, sphygmomanometers and clinical hammers, for all patients in the ward without disinfecting them between patients. Nebulizers were used 14 times and no disinfectant was observed before or after use. A doctor explained,
“The same oxygen tube is used for almost all the patients. There is only one oxygen cylinder and one mask in the ward and that mask is used for every patient.”
Thermometers, used in the mouth, were partially dipped in bottles half-filled with disinfectant after each patient use.
A total of 3,373 handwashing opportunities were noted during 48 hours, and occurred before or after family caregivers and healthcare workers cared for patients, after ward occupants blew their noses, coughed, sneezed, or vomited, and before ward occupants consumed food (
Handwashing opportunities | Frequency |
During patient care | |
552 | |
550 | |
458 | |
389 | |
164 | |
144 | |
88 | |
74 | |
64 | |
61 | |
56 | |
46 | |
40 | |
22 | |
After blowing nose, coughing, sneezing and vomiting | |
79 | |
56 | |
20 | |
4 | |
Before eating | 506 |
Total | 3373 |
No staff mentioned knowledge of any policy or written rule on infection control. All categories of staff reported inadequate supplies of cleaning and disinfection products, bed sheets, soap and hand sanitizer. Support staff reported,
“They gave us three Harpics [a brand of toilet cleaning product] last month. We must use Harpic for professors’ handwashing station daily. There are also doctors’ handwashing stations and toilets which have to be cleaned. Two of three Harpics are used for cleaning these toilets. We have to manage cleaning the five to six patient toilets with the remaining one.”
There was no autoclave in the wards and nurses reported disinfecting medical instruments by boiling, which the team also observed, or immersing in chlorine-water solution for 10 minutes. A doctor stated,
“An instrument should be boiled for at least 30 minutes. We are so loaded with patients that nurses are only dipping instruments in hot water.”
Nurses and support staff reported using antiseptic liquid or saline to wash hands on some occasions when soap was not available. Nurses reported using surgical masks while making patient beds, which the team never observed. Nurses also reported that they could change bed sheets only every one to two weeks due to inadequate supply. They mentioned family caregivers sometimes took linens soiled with patient body secretions and excretions to wash at home because water was often unavailable in the hospital. Support staff reported using gloves when caring for patients with infectious diseases like hepatitis B, but the team never observed this. Staff also reported that supplies could not be accessed in the evening or at night when the nurse-in-charge was not on ward duty.
Overcrowding, inadequate sanitary facilities, lack of routine cleaning, lack of basic infection control measures and improper waste management combined to create numerous opportunities for transmission of infection in the study wards. This environment posed a threat of infection to all ward occupants.
Crowding in hospitals facilitates the spread of many diseases
Hospital surfaces could be potential reservoirs of nosocomial pathogens that can survive for a few days to several months
Most hospital infections are acquired via direct contact
The presence of animals inside wards poses a threat of transmission of zoonotic diseases. In a geriatric care center, a cat was associated with an outbreak of epidemic MRSA
Improper handling and unsafe disposal of hospital waste is a public health concern globally
This study was conducted in only three public tertiary hospitals that were not randomly chosen; therefore the findings cannot be generalized to other government tertiary hospitals, private clinics or other non-government hospitals in Bangladesh. However, our findings are consistent with other studies in Bangladeshi tertiary hospitals that have reported crowding in wards
Public hospitals play a crucial role in ensuring healthcare services for the poor in Bangladesh
icddr, b acknowledges the commitment of the Government of the People’s Republic of Bangladesh (GoB) to our research efforts with gratitude. We acknowledge, with gratitude, the cooperation of the hospital faculty, administrators and collaborators for their support of the study. We are indebted to the study participants for their valuable time and information. We thank Dorothy Southern and Meghan Scott for their support in reviewing and editing this manuscript and Md. Jaynal Abedin for his contribution in the calculation of the quantitative data.