Effect of Mandated Nurse–Patient Ratios on Patient Wait Time and Care Time in the Emergency Department


  • Presented at the Society for Academic Emergency Medicine annual meeting, New Orleans, LA, May 2009.

  • This study was funded by the 2007 Emergency Medicine Foundation/Emergency Nursing Association Foundation Team Grant/ED Overcrowding Research Award.

Address for correspondence and reprints: Theodore Chan, MD; e-mail: tcchan@ucsd.edu.


Objectives:  The objective was to evaluate the effect of mandated nurse–patient ratios (NPRs) on emergency department (ED) patient flow.

Methods:  Two institutions implemented an electronic tracking system embedded within the electronic medical record (EMR) of two EDs (an academic urban, teaching medical center—Hospital A; and a suburban community hospital—Hospital B), with a combined census of 60,000/year, to monitor real-time NPRs and patient acuity, such that compliance with state-mandated ratios could be prospectively monitored. Data were queried for a 1-year period after implementation and included patient wait times (WTs), ED care time (EDCT), patient acuity, ED census, and NPR status for each nurse, patient, and the ED overall. Median WT and EDCT with interquartile ranges (IQRs) were analyzed to determine the effect of NPR status of each patient, nurse, and the ED overall. To control for factors that could affect the “within the mandated ratio” and the “outside of the mandated ratio” status, including patient volume and acuity, log-linear regression models were used controlling for specified factors for each hospital facility and combined.

Results:  There were a total of 30,404 (50.9%) patients who waited in the waiting room prior to being placed in an ED bed (53.8% at Hospital A and 46.4% at Hospital B). Patients who waited at Hospital A waited a median duration of 55 minutes (IQR = 15–128 minutes), compared with 32 minutes (IQR = 12–67 minutes) at Hospital B with a combined median WT of 44 minutes (IQR = 13–101 minutes). In the log-linear regression analysis, WTs were 17% (95% confidence interval [CI] = 10% to 25%, p < 0.001) longer at Hospital A and 13% (95% CI = 3% to 24%, p = 0.008) longer at Hospital B (combined 16% [95% CI = 10% to 22%, p < 0.001] longer at both sites) when the ED overall was out-of-ratio compared to in-ratio. There were a total of 45,660 patients discharged from both EDs during the study period, from which EDCT data were collected (26,894 in Hospital A and 18,766 in Hospital B). Median EDCT was 184 minutes (IQR = 97–311 minutes) at Hospital A, compared to 120 minutes (IQR = 63–208 minutes) at Hospital B, for a combined median EDCT of 153 minutes (IQR = 81–269 minutes). In the log-linear regression analysis, the EDCT for patients whose nurse was out-of-ratio were 34% (95% CI = 30% to 38%, p < 0.001) longer at Hospital A and 42% (95% CI = 37% to 48%, p < 0.001) longer at Hospital B (combined 37% [95% CI = 34% to 41%, p < 0.001] longer at both sites) when compared to patients whose nurse was in-ratio.

Conclusions:  In these two EDs, throughput measures of WT and EDCT were shorter when the ED nurse staffing were within state-mandated levels, after controlling for ED census and patient acuity.

ACADEMIC EMERGENCY MEDICINE 2010; 17:545–552 © 2010 by the Society for Academic Emergency Medicine

Nurses play an essential role in patient care in acute care hospitals. Over the past decade, there has been growing concern that changes in nurse staffing and increased patient workloads are threatening the quality of patient care.1 Mandated nurse–patient ratios (NPRs) have been proposed to address perceived inadequate nurse staffing levels in acute care hospitals. In 2004, California enacted minimum NPRs following the passage of Assembly Bill 394 of the Safe Staffing Law in 1999. This law set forth minimum staffing ratios throughout the hospital, including the emergency department (ED), where the minimum ratio is set at one nurse for every four patients, with lower ratios for critical care patients (one nurse for every two patients) and critical trauma patients (one nurse for every one patient).2 Similar minimum NPR staffing regulations and laws have been proposed in other states throughout the country, as well as at the federal level in the U.S. Congress.3

In addition to direct patient care, nurse staffing is a critical component for overall ED operations, efficiency, and patient flow. Minimum staffing regulations may be particularly challenging for EDs because patient census, load, and acuity can fluctuate minute-by-minute. As a result, most California EDs have had periods of time out of compliance with the state-mandated NPRs.4 Opponents argue that minimum NPRs in the ED may negatively affect patient throughput by reducing available patient care areas that cannot be staffed to meet the ratio requirement.5 Conversely, proponents believe NPRs reduce nurse workloads to manageable levels such that patient care, evaluation, and disposition tasks can be completed in a timely manner, thus improving ED patient flow and throughput. We sought to examine the effect of mandated NPRs on two EDs, an academic, urban ED and Level 1 trauma center and a suburban, community ED, to determine the effect of the statutory change on ED throughput and patient flow.


Study Design

We conducted a multicenter, prospective observational comparison study to assess the effect of mandated NPRs on ED patient flow. Our hypothesis was that the NPR status based on state mandates does influence our two primary outcomes measures, patient wait time (WT) and ED care time (EDCT). This project was approved and informed consent requirements were waived by our university’s institutional review board.

Study Setting and Population

The study was conducted at two EDs (providing care for adult and pediatric patients) where state-mandated NPRs have been enacted for all acute care hospital settings, including the ED. One site (Hospital A) was an urban, academic teaching hospital (Level 1 trauma center) 24-bed ED (including a four-bed urgent care “fast-track” area) with an annual census of approximately 37,000 visits. The other site (Hospital B) was a suburban community hospital, with a 15-bed ED (including a four-bed urgent care “fast-track” area), with an annual census of approximately 23,000 visits. The EDs are staffed with nurses based on anticipated day-of-week and time-of-day patient arrivals and, thus, not all beds are staffed 24 hours a day, 7 days a week. There is a designated triage and charge nurse with no patient care assignments at each site. In addition, staff may call in sick or fail to show for an expected shift. Thus, patient load can exceed both staffing and beds at times, and both EDs utilize hallway beds for overflow when needed.

Study Protocol

The study was conducted over a 1-year period (January 1, 2008, through December 31, 2008) after enactment of the state-mandated NPR levels. Data collected during the study period included patient WT, EDCT, ED census (including all patients in the ED and waiting room), patient triage acuity, acuity as determined throughout the patient care time by the nurse, patient disposition, and NPR status every 10 minutes for each ED nurse, for each ED patient, and for the ED as a whole. WT was defined as time from triage to placement in an ED bed, and EDCT was defined as time from placement in an ED bed to either discharge from the ED or transfer to an inpatient bed.

To collect patient acuity and staffing ratio measurements, we implemented an automated electronic tracking system imbedded within the electronic medical record (EMR) of both sites to monitor real-time NPR status in the ED. In both EDs, all patient tracking, as well as nursing–patient assignments and clinical documentation for patient visits, is entered electronically in real time into the EMR. Nurses are assigned specific beds in the ED, and patients are placed in those beds as they arrive. The system tracked ED nurse–patient assignments and specific patient acuity such that compliance with state-mandated ratios could be recorded for the ED at any time. Patient assignments for each nurse were recorded from the electronic ED status tracking board. Patient acuity was tracked and recorded by the ED nurse with every nurse assessment and note entry into the EMR. With each entry, the nurse is required to select an acuity level to confirm or record any changes in patient acuity. Records with the highest likelihood of requiring a patient acuity change during treatment, such as discordance between triage acuity and discharge status, were reviewed periodically by a research nurse to ensure compliance and accuracy of acuity determination (i.e., a patient with an emergent triage level, but who was discharged to home).

Because both the number of patients assigned to a given nurse and the acuity of patients could change during their course in the ED, NPR data were extracted and recorded from the EMR of both EDs every 10 minutes during the 1-year study period. This NPR data included whether a specific nurse’s patient assignment was within state-mandated maximum ratios based on number and acuity (1:1 for trauma resuscitation patients, 1:2 for critical patients, 1:4 for all other ED patients) at any given point in time during patient care. In-ratio or out-of-ratio status was then determined for each ED nurse, each ED patient, and the ED overall as a whole. For each ED nurse, the in-ratio or out-of-ratio status was based on his or her patient assignments (including the tracking of breaks and coverages) and the acuity of those patients at a given time. Nurses were considered out-of-ratio if their patient load exceeded state regulations for more than 20 minutes of patient care time. For each ED patient, in-ratio or out-of-ratio status was based on the status of the nurse assigned to care for him or her in the ED at that time. If the nurse was out-of-ratio at a given time, then all patients cared for by that nurse were in out-of-ratio status at that time. For overall ED status, in-ratio was defined as having all nurses in the ED being in-ratio at that time and out-of-ratio as having any nurse out-of-ratio at that time.

For ED WT, each patient who waited to get to an ED bed was stratified as in-ratio or out-of-ratio based on whether the ED as a whole was in-ratio or out-of-ratio at the time that patient initially went to the waiting room. The ED was considered in-ratio if all staff nurse–patient assignments were within mandated NPR levels and considered out-of-ratio if any staff nurses were beyond those levels. This stratification was chosen because unlike EDCT, where nursing care may have a direct effect on care time, WT is more likely to be affected by the ED staffing prior to actual care, that is, during the time the patient was actually waiting to get to an ED bed, rather than the staffing status once he or she was in an ED bed. For EDCT, each ED patient discharged from the ED was stratified based on whether he or she had more than 20 minutes of total patient care time provided by a nurse who was in excess of state-mandated NPR levels. We believe that this amount of time out-of-ratio would potentially affect patient flow, as opposed to shorter periods of time where the nurse may be in the process of reassigning patients to address NPR levels. We excluded admitted patients from the analysis of EDCT because of the potential effect of boarding on care time duration in the ED. However, patients who were admitted, but boarding in the ED and cared for by an ED nurse, were still included in the determination of the nurse’s NPR status.

An example scenario would be as follows: a nurse is assigned to four standard patients in the ED and is within-ratio. No other nurse is out-of-ratio in terms of patient assignments. At this time, the nurse, all four patients and the ED are considered in-ratio. One of the patients becomes critically ill requiring intubation, invasive monitoring, and admission to the intensive care unit. The nurse continues to care for this patient, as well as the other three patients, for 30 minutes before the patients are reassigned to other nurses. During this time, the nurse is considered out-of-ratio (with one critical and three regular ED patients). Each of the four patients is considered out-of-ratio because the nurse caring for them has a patient load exceeding the regulations. The ED as a whole is also considered out-of-ratio because one of the nurses is at that status during this time.

Data Analysis

Data were analyzed to determine if NPR status for each patient, nurse, or ED classified as in-ratio or out-of-ratio affected ED flow parameters of WT and patient EDCT after controlling for hospital, census, and three-tiered initial patient triage acuity level (emergent, acute, urgent).

Descriptive and demographic patient statistics are reported as frequencies and percentages for categorical variables. Age was normally distributed and reported as means with standard deviations (±SDs). WT and EDCT are reported as median times with interquartile ranges (IQRs) because of the skewed nature of the data. Patients who left without treatment (LWOT) were excluded from WT calculations because of inaccurate data. Overall differences by hospital were assessed using a chi-square test for categorical variables and a t-test for age. Because ratio status (in or out) is likely influenced by factors associated with ED volume (such as patient census and acuity), log-linear regression models, using the natural log of WT and EDCT due to their skewed properties, were used to assess the effect of NPRs while controlling for census at time of waiting (for WT), ED bed placement (for EDCT), patient acuity level at triage, and hospital facility and were repeated stratifying by facility. The models estimate the difference in the outcome measures (WT or EDCT) between in-ratio and out-of-ratio status, while holding the controlling factors associated with ED volume constant. For the models, the natural-log-transformed WT or EDCT was used as the dependent variable, and the other factors were included as covariates. The percentage change in WT and EDCT as a result of the change in ratio status, while controlling for the other factors, is reported from the models. Statistical analyses were performed using SPSS 17.0 (SPSS Inc., Chicago, IL).


During the 1-year study period, the combined census of both EDs was 59,733 visits. Patient population characteristics are presented in Table 1. Overall, the mean (±SD) patient age was 44.2 (±19.8) years, with a near equal distribution between men and women. The majority of patients were treated and discharged from the ED (n = 45,660; 76.4%). Admissions from the ED accounted for the next highest disposition (n = 10,422; 17.4%). A higher proportion of patients were male, waited for ED services, and had a higher acuity at Hospital A compared to Hospital B. At Hospital A, the rates of patients who were seen and discharged were lower and those who LWOT were higher.

Table 1. 
Population Characteristics of Patients Seen in the Two Participating EDs, January 1, 2008, Through December 31, 2008
CharacteristicHospital A
(n = 36,462)
Hospital B
(= 23,271)
Both Facilities
(N = 59,733)
  1. *Significant difference between Hospital A and Hospital B (p<0.05).

Age (yr),* mean (±SD)43.9 (±18.3)44.9 (±21.9)44.3 (±19.8)
Sex*Number (%)Number (%)Number (%)
 Male19,902 (54.6)10,181 (43.7)30,083 (50.4)
 Female16,560 (45.4)13,090 (56.3)29,650 (49.6)
 Waited for services*19,608 (53.8)10,796 (46.4)30,404 (50.9)
Triage acuity*
 Emergent4,139 (11.4)1,421 (6.1)5,560 (9.3)
 Acute21,815 (59.8)12,563 (54.0)34,378 (57.6)
 Urgent10,508 (28.8)9,287 (39.9)19,795 (33.1)
 Admitted to hospital6,869 (18.8)3,553 (15.3)10,422 (17.4)
 Left against medical advice529 (1.5)125 (0.5)654 (1.1)
 Discharged from ED26,894 (73.8)18,766 (80.6)45,660 (76.4)
 Expired in the ED48 (0.1)5 (0.0)53 (0.1)
 Left without treatment1,880 (5.1)422 (1.9)2,302 (3.9)
 Transfer to another facility242 (0.7)400 (1.7)642 (1.1)


There were a total of 30,404 (50.9%) patients who waited in the waiting room for ED services in both hospitals prior to being placed in an ED bed (53.8% at Hospital A and 46.4% at Hospital B). Characteristics of patients who waited are reported in Table 2. Patients who waited at Hospital A waited a median duration of 55 minutes (IQR = 15–128 minutes), compared with 32 minutes (IQR = 12–67 minutes) at Hospital B, with a combined median WT of 44 minutes (IQR = 13–101 minutes). Overall and at each facility, patients triaged at the highest emergent acuity level waited a shorter period of time than those triaged at the lower acuity levels.

Table 2. 
Median WTs and IQRs by Characteristics of Patients Who Waited for ED Services, January 1, 2008, Through December 31, 2008
CharacteristicHospital AHospital BBoth Facilities
Number (%)Median WT, Minutes (IQR)Number (%)Median WT, Minutes (IQR)Number (%)Median WT, Minutes (IQR)
  1. IQR = interquartile range; LWOT = left without treatment; WT = wait time.

  2. *WT calculations exclude LWOT patients.

  3. †Statistically significant difference proportions between Hospital A and Hospital B (p < 0.05).

  4. ‡At time patient went to waiting room.

Waited*†19,608 (53.8)55 (15–128)10,796 (46.4)32 (12–67)30,404 (50.9)44 (13–101)
Triage acuity†
 Emergent1,081 (5.6)17 (6–48)181 (1.7)9 (3–27)1,262 (4.2)15 (5–45)
 Acute9,897 (51.3)85 (27–172)4,953 (46.7)36 (13–77)14,850 (49.7)62 (19–138)
 Urgent8,328 (43.1)39 (11–88)5,463 (51.6)30 (11–61)13,791 (46.1)34 (11–75)
ED ratio status†‡
 In-ratio17,116 (88.7)52 (14–124)9,805 (92.5)31 (11–66)26,921 (90.0)42 (13–98)
 Out-of-ratio2,190 (11.3)76 (24–154)792 (7.5)45 (20–78)2,982 (10.0)63 (22–132)

In terms of nurse ratio status, 10% (2,982) of all patients who waited for ED services went to the waiting room at a time when the ED overall was at out-of-ratio status. WT time was longer for these patients, compared to those who went to the waiting room at a time when the ED was in-ratio: median WT was 42 minutes (IQR = 13–98 minutes) when the ED was in-ratio at the start of the patient’s WT, compared to 63 minutes (IQR = 22–132 minutes) when the ED was out-of-ratio at that time. In the log-linear regression analysis, after controlling for ED census, patient triage acuity, and hospital, ED WTs were 16% (95% confidence interval [CI] = 10% to 22%, p < 0.001) longer for patients who waited while the ED was out-of-ratio, compared to in-ratio. When stratified by facility, the WT 17% (95% CI = 10% to 25%, p < 0.001) was longer at Hospital A, and 13% (95% CI = 3% to 24%, p = 0.008) longer at Hospital B when the ED overall was out-of-ratio compared to in-ratio at each respective hospital.


There were a total of 45,660 patients discharged from both EDs during the study period, from which EDCT data were collected (26,894 in Hospital A and 18,766 in Hospital B). Characteristics of treated and discharged patients are reported in Table 3. Median EDCT was 184 minutes (IQR = 97–311 minutes) at Hospital A, compared to 120 minutes (IQR = 63–208 minutes) at Hospital B, for a combined median EDCT of 153 minutes (IQR = 81–269 minutes). Overall, patients triaged at the emergent acuity level had the longest median EDCT (295 minutes, IQR = 196–427), compared to those triaged at an acute level (218 minutes, IQR = 138–331 minutes) or urgent level (83 minutes, IQR = 51–130 minutes), and this was consistent at both facilities.

Table 3. 
Median EDCTs and IQRs by Characteristics of Patients Who Were Seen and Discharged, January 1, 2008, Through December 31, 2008
CharacteristicHospital AHospital BBoth Facilities
Number (%)Median EDCT, Minutes (IQR)Number (%)Median EDCT, Minutes (IQR)Number (%)Median EDCT, Minutes (IQR)
  1. EDCT = ED care time; IQR = interquartile range.

  2. *Significant difference proportions between Hospital A and Hospital B (p < 0.05).

Seen and discharged*26,894 (73.8)184 (97–311)18,766 (80.6)120 (63–208)45,660 (76.4)153 (8–269)
Triage acuity*
 Emergent1,804 (6.7)318 (212–460)665 (3.5)244 (160–359)2,469 (5.4)295 (196–427)
 Acute15,415 (57.3)244 (159–365)9,172 (48.9)178 (112–270)24,587 (53.8)218 (138–331)
 Urgent9,675 (36.0)91 (60–143)8,929 (47.6)73 (42–117)18,604 (40.7)83 (51–130)
Nurse ratio status
 In-ratio24,949 (92.8)179 (95–302)17,356 (92.5)116 (61–202)42,305 (92.7)149 (79–261)
 Out-of-ratio1,945 (7.2)261 (150–429)1,410 (7.5)165 (91–287)3,355 (7.3)225 (117–367)

In terms of nurse ratio status, a total of 3,355 discharged patients (7.3%) were treated by an ED nurse who was out-of-ratio for >20 minutes during the patient’s care time. Median EDCT was longer for these patients (225 minutes, IQR = 117–367 minutes) compared to those patients whose ED nurse remained in-ratio during the patient’s care time (149 minutes, IQR = 79–261 minutes). In the log-linear regression analysis, after controlling for ED census, patient triage acuity, and hospital, the EDCTs for patients whose nurse was out-of-ratio were 37% (95% CI = 34% to 41%, p < 0.001) longer than those patients whose nurse was in-ratio. When stratified by facility, the EDCT for patients whose nurse was out-of-ratio was 34% (95% CI = 30% to 38%, p < 0.001) longer at Hospital A and 42% (95% CI = 37% to 48%, p < 0.001) longer at Hospital B, compared to patients whose nurse was in-ratio at each respective hospital.


Over the past two decades, nursing workforce shortages, financial pressures on health care, and advances in medical care have resulted in marked changes in acute care nursing. Growing concern over nursing workload on quality of care, error rates, and outcomes has led to increased focus on nurse staffing levels in acute care hospital settings. Many states have enacted legislation requiring hospitals to develop and implement nurse staffing plans and to even publicly report their staffing levels.6 The most aggressive of these types of measures are proposals to statutorily establish minimum nurse–patient staffing levels in health care settings. These minimal staffing, NPR regulations have generated significant controversy among providers, administrators, patient advocates, and professional organizations and unions.3,7 In its 1996 report To Err is Human, the Institute of Medicine reported that the literature on NPRs on safety and quality of care was inadequate.8 Since that time, a number of studies have reported an association between acute care staffing levels and improved patient outcomes.1,9–11 However, others argue that there is a lack of definitive data to show a direct causal connection for improved safety and patient outcomes.7

Staffing ratio regulations have been proposed in state legislatures in over 14 states, as well as at the federal level in the U.S. Congress.6 In 2004, California became the first state in the U.S. to enact mandated NPR staffing levels for hospitals, including the ED, and now has become a model for some in support of similar regulations in other states.12 Studies evaluating the effect of the new law indicate that significant changes in nurse staffing, as well as an overall increase in total nurse patient care hours, have occurred after implementation of the law.13,14 The effect of the law on patient safety and quality has been less clear. Studies comparing metrics sensitive to nursing care, such as patient falls and decubitus ulcer rates, have shown little change before and after the law’s enactment.12,13 Moreover, hospital patient flow, as measured by average length of inpatient stay, also has not changed after the ratios were implemented.12

Even less well studied has been the effect of the state-mandated NPR staffing levels on the ED, where patient–nurse ratio levels have been perceived as a marker for ED crowding and potentially compromised care.15 Initially, concerns were raised that mandated ratios, in combination with the ongoing shortage in nursing workforce, would reduce available staffed hospital inpatient and ED beds, potentially reducing access to care for many.5 Indeed, anecdotal reports suggested that patient WTs and ambulance diversion increased as a result of admitted patients boarding in the ED due to the lack of hospital nurse staffing.12

A recent study by Weichenthal and Hendey16 reported that ED WT, patient ED throughput, and admission time all increased at a single California hospital ED when comparing the year prior to the year after mandated ratios were enacted. Other measures, such as time to antibiotics for pneumonia patients and left without being seen rates, however, actually improved after implementation of ratios. A general weakness of these comparative studies on the California law is their pre/post study design, which compare measures before and after the ratio law enactment. As a result, a number of potential confounding variables are introduced that may have affected results unrelated to the nurse ratio law (such as hospital initiatives to improve timeliness of antibiotic administration). Moreover, because the ratio law was initially passed in 1999, but not enacted until 2004, many hospitals began increasing nursing staff in various units before the law was implemented. Thus, it is unclear the degree to which actual NPRs actually changed from the end of 2003 to the beginning of 2004.

Because of fluctuating patient census and acuity, the ED is a unique setting whereby NPR levels vary over time, often going above and below mandated levels. The ED remains a very dynamic environment with a mandate to screen every patient, as opposed to the inpatient setting, which can limit patient placement to staffed beds. As a result, the ED provides a natural experiment to compare times when in-ratio and out-of-ratio in this setting. To conduct such a study, we leveraged informatics tools to develop real-time tracking software embedded in our ED’s operational EMR to record and study nurse patient assignments, patient acuity and any changes in acuity, and overall ED staffing status. Doing so allowed us to track ratio status by patient, nurse, and the ED overall.

In our study, approximately 1 of every 13 patients seen and discharged was cared for by a nurse who was out-of-ratio for a significant portion of the patient’s EDCT. We did not specifically track the underlying reasons for out-of-ratio status, as this was beyond the scope of our study. However, anecdotally, nurses could have exceeded NPR staffing regulations in a variety of scenarios, including when a patient deteriorated and became more critically ill while in the ED (thus requiring a higher level of nurse staffing), excessive numbers of patients arrived either walk-in or by ambulance and were taken into the ED care areas, or staff shortages took place (i.e., sick calls, failures to show for assigned shifts). In the study EDs, interventions to address out-of-ratio periods included redistributing patient assignments to the nursing staff, using administrative nurse staff for patient care (i.e., to address nurse staff shortages temporarily), or having patients wait longer for care.

It is important to note that the mandated ratios do not include or factor in waiting room patients. Also, the determination of “critical” is somewhat vague in the state regulations and subject to some level of interpretation. In our two study EDs, “critical” was determined by the need for intubation or advanced airway management (e.g., continuous positive airway pressure), need for invasive monitoring (e.g., arterial line or central venous monitoring such as SVO2 or pulmonary artery pressure monitoring), or admission to an intensive care unit setting.

In our combined facility log-linear regression modeling, we found that EDCT was, on average, 32% longer when the nurse was out-of-ratio, and WT increased by 10% for those waiting when the ED was out-of-ratio, while controlling for triage acuity, ED census, and facility. This equates to an increase of over an hour to a normally three hour EDCT and of about 10 minutes to a 60-minute WT when out-of-ratio. On the face of it, these results are not surprising in that when the ED is busier, it is more likely staff would exceed mandated ratios and patient care processes would take longer, as demonstrated in WT and EDCT. These findings are reflective of Little’s Law in Operations Research, which states in essence that in stable systems over the long term, the average number of patients is a function of both the average arrival rate and the average length of stay.17

To address this issue in our analysis, we controlled for both ED volume and patient acuity when comparing outcome measures between in-ratio and out-of-ratio groups. Even after controlling for these factors when the ED is more likely to exceed mandated NPRs, we found longer WTs and EDCTs. This finding was seen combined and separately at both hospital EDs studied, representing two very different sites with diverse patient populations and census, as well as differing baseline patient flow parameters and times.

Our study results indicate that efforts to staff EDs within mandated NPR levels do have a beneficial effect on patient flow. This again may simply reflect Little’s Law, as it is unlikely that controlling just for census and acuity would account for all the factors affecting how busy an ED may be. However, our findings do suggest that efforts to have more patients cared for in the ED by nurses who are out-of-ratio in an effort to increase patient throughput may be less efficacious than one would hope.


Our study was conducted over a relatively short period of time (1 year) at only two EDs in California. While our sites were quite different (urban academic trauma center and suburban community hospital), our findings may not be directly applicable to other EDs in different settings. In addition, it is possible that other initiatives undertaken during the study period and unrelated to nurse staffing may have influenced our results. However, during the study period there were no such major operational initiatives focused on patient flow initiated in our study EDs.

Our data relied on the nursing staff to determine and record patient acuity accurately and in a timely manner into our EMR system. It is possible that inaccurate data reporting, lack of timely recording (which would affect our timed data collection), and unknown biases of the staff may have affected our results. However, we sampled data for quality review purposes and found no instances where the treatment acuity level or changes in level were not appropriate based on the nurse and physician notes. It is also possible that there were inaccuracies in other time measures. However, the data were queried from an EMR, and there were no measures that were considered significant outliers.

We selected a threshold of 20 minutes of patient care time above the mandated NPR level to be considered out-of-ratio. We believe this amount of out-of-ratio care time would reveal any effect on patient flow parameters, should such an effect have occurred. However, it is possible that a longer period of out-of-ratio care time might have been necessary to demonstrate an effect that we could not detect. Similarly, analysis of WT was based on a static assessment of ED status, with a threshold of one or more nurses being out-of-ratio at that time, although other nurses might have been in-ratio at the same time, and the status of the ED might change during the patient’s WT. Determination of a proportion of care time in-ratio or out-of-ratio may have been a more rigorous approach rather than relying on a dichotomized set threshold time. However, such an approach would still rely on the threshold level of the state regulations regarding NPR staffing levels.

Because we wanted to evaluate the impact of California’s state-mandated levels, we classified in-ratio and out-of-ratio status based on those regulations, with 1 to 4 as the maximum NPR and lower ratios depending on patient acuity. It is possible that at this staffing level, being in-ratio or out-of-ratio may make little difference; that is, there may be only an incremental change between a 1 to 4 ratio considered in-ratio and a 1 to 5 level considered out-of-ratio. Moreover, these levels may be significantly lower than typical ED nurse staffing ratios elsewhere and again not applicable to other settings. In addition, we were unable to collect the specific reason a nurse may have been out-of-ratio (e.g., change in patient acuity or increased numbers of patients), which would have provided more detailed information to analyze the effect of the ratio standard.

When EDs are busy (because of volume, acuity, or both), there is a greater likelihood the ED would be out-of-ratio and that all patient tasks (such as ancillary testing) would take longer to perform regardless of staffing. While we used regression modeling in our analysis to control for markers of how busy the ED was at a given point in time, doing so may not have sufficiently accounted for all the other factors that can affect WT and length of stay. For example, we did not control for factors such as consult and ancillary testing delays that can affect EDCT for example.

Our study focused solely on nurse staffing NPRs in the ED. We did not assess the effect of mandated NPR levels in the rest of the hospital. Inpatient NPR levels and mandates likely affect ED patient flow by restricting staffed hospital beds for admitted patients, worsening the problem of boarding in the ED, a recognized major contributor to ED crowding.12,18 While we restricted our study to patients who waited or were discharged from the ED, we have previously shown that patients boarding in the ED can affect patient flow, WT, and EDCT for discharged patients.19 Finally, while we evaluated the effect of mandated NPRs in the ED on patient flow parameters, we did not evaluate specific clinical quality of care measures or overall patient outcomes.


In this study of two California EDs following implementation of state-mandated nurse–patient ratio levels, ED throughput measures of wait time and ED care time were shorter when the ED nurse staffing was within mandated levels, after controlling for ED census and patient acuity. Further study is needed to assess the effect of nurse–patient ratio mandates on patient clinical outcome measures.