Population Health And Safety - USMLE Step 3

SECTION 1

Historical Context & Motivation

The concept of population health emerged from the recognition that disease and safety are not merely individual phenomena but are shaped by broad social, environmental, and systemic forces. Long before germ theory, physicians and public officials observed that illness clustered in specific communities and that certain environmental conditions—contaminated water, overcrowding, and poor sanitation—produced predictable patterns of morbidity and mortality. The evolution from treating individual patients to safeguarding entire populations represents one of the most consequential paradigm shifts in the history of medicine, and it underpins much of what modern primary care physicians encounter on the USMLE Step 3.

1854 • John Snow & the Broad Street Pump

John Snow's investigation of the London cholera outbreak established the epidemiological method by mapping cases and tracing them to a contaminated water source, demonstrating that population-level data could identify disease causation before laboratory confirmation.

1906 • Pure Food and Drug Act

The United States enacted its first major federal legislation governing food and pharmaceutical safety, establishing the principle that government agencies bear responsibility for protecting population health through regulation.

1948 • Framingham Heart Study Launch

This landmark prospective cohort study began following over 5,000 residents of Framingham, Massachusetts, ultimately identifying major cardiovascular risk factors—hypertension, hypercholesterolemia, smoking—and solidifying the evidence base for population-level prevention.

1999 • IOM Report: To Err Is Human

The Institute of Medicine estimated that 44,000–98,000 Americans died annually from preventable medical errors, catalyzing a national focus on patient safety as a population health concern and leading to systems-based safety reforms.

2010 • Affordable Care Act & Accountable Care

The ACA established Accountable Care Organizations (ACOs) and value-based payment models, formally linking physician reimbursement to population health outcomes such as preventive screening rates, hospital readmission reduction, and chronic disease management.

These milestones illustrate a central question that Population Health and Safety addresses: how can clinicians, institutions, and policy frameworks work together to prevent disease, reduce harm, and improve health outcomes for defined groups of people rather than one patient at a time? The USMLE Step 3 tests your ability to apply these principles in clinical and systems-based contexts, making this topic essential for every graduating physician.

SECTION 2

Core Principles & Definitions

Population health integrates clinical medicine with epidemiology, health policy, and safety science. To approach Step 3 questions confidently, you must internalize several foundational concepts that recur across clinical vignettes. These principles connect screening guidelines, quality improvement metrics, and patient safety interventions to a unified framework for population-level thinking.

Levels of Prevention

Primary prevention prevents disease onset (e.g., vaccination). Secondary prevention detects disease early (e.g., mammography screening). Tertiary prevention reduces disability from established disease (e.g., cardiac rehabilitation).

Social Determinants of Health

Non-medical factors—income, education, housing, food access, and neighborhood safety—account for an estimated 30–55% of health outcomes. Addressing these determinants is critical for reducing health disparities across populations.

Screening Criteria (Wilson & Jungner)

Effective population-based screening requires that the condition is an important health problem, a suitable test exists, early treatment is more effective than late treatment, and the program is cost-effective relative to the overall healthcare expenditure.

Patient Safety & Systems Thinking

The Swiss Cheese Model (James Reason) illustrates that medical errors result from alignment of multiple system failures, not single-person negligence. Safety cultures emphasize reporting, root cause analysis, and blame-free improvement.

Quality Measures & Reporting

Organizations such as CMS, NCQA, and the Joint Commission use standardized metrics—HEDIS measures, Hospital Compare, and Physician Quality Reporting System (PQRS)—to track population-level health outcomes and incentivize quality improvement.

✦

KEY TAKEAWAY

Think of population health like municipal water treatment: instead of giving every household an individual purification device (individual medicine), you treat the water supply at its source so that everyone benefits simultaneously. Screening programs, vaccination campaigns, and safety systems are the 'treatment plants' of healthcare—they address risk at the systems level before disease manifests in individual patients.

SECTION 3

Visual Framework: The Population Health Pyramid

The population health impact pyramid, adapted from Thomas Frieden's model, organizes interventions by their reach and the degree of individual effort required. Interventions at the base of the pyramid—such as socioeconomic improvements and changes to default conditions—affect the largest number of people with the least individual effort, while those at the apex—clinical interventions and individual counseling—reach fewer individuals and demand more personal engagement. Understanding this hierarchy helps clinicians identify where their recommendations fit and why policy-level changes often produce the greatest population-level health gains.

The Frieden pyramid demonstrates that socioeconomic interventions and changes to default conditions (base layers) produce the greatest population health impact, while individual counseling and clinical interventions (apex) require more effort per person and reach fewer individuals. USMLE Step 3 questions often ask you to identify which level of intervention is most appropriate for a given scenario.

As you ascend the pyramid, each tier represents a progressively more targeted and individually intensive intervention. Water fluoridation, for example, sits in the second tier because it changes the default environment for an entire community without requiring any individual behavior change. In contrast, prescribing a statin for cardiovascular disease risk reduction is a clinical intervention that requires patient adherence, monitoring, and follow-up. The most efficient population health strategies operate at the base, yet physicians must understand all tiers because Step 3 questions frequently test the distinction between individual-level clinical care and systems-level population health actions.

SECTION 4

Key Metrics & Quantitative Framework

Population health relies on quantitative measures to assess disease burden, evaluate interventions, and guide resource allocation. Mastery of these metrics is essential for the USMLE Step 3, where clinical vignettes may present epidemiological data and expect you to calculate or interpret key values. The following equations represent the core quantitative tools that population health professionals and primary care physicians use to make evidence-based decisions.

INCIDENCE RATE

Incidence Rate = (Number of new cases during a time period) ÷ (Population at risk during that period) × 10ⁿ

Incidence measures the rate of new disease occurrence. The multiplier 10ⁿ (often 1,000 or 100,000) converts the rate to a manageable number. Incidence is used to assess disease risk and evaluate the effectiveness of preventive interventions.

PREVALENCE

Prevalence = (Number of existing cases at a point in time) ÷ (Total population at that point) × 100

Prevalence captures the total burden of disease in a population at a specific time. Point prevalence is influenced by both incidence and disease duration. A disease with low incidence but long duration (e.g., diabetes) may have high prevalence.

NUMBER NEEDED TO TREAT (NNT)

NNT = 1 ÷ (Absolute Risk Reduction) = 1 ÷ (CER − EER)

Where CER = control event rate and EER = experimental event rate. NNT tells you how many patients you need to treat with an intervention for one additional patient to benefit. Lower NNT values indicate more effective interventions. This is a high-yield metric for Step 3.

SENSITIVITY AND SPECIFICITY

Sensitivity = TP ÷ (TP + FN) | Specificity = TN ÷ (TN + FP)

Where TP = true positives, FN = false negatives, TN = true negatives, and FP = false positives. Sensitivity (SnNOut) rules out disease when negative. Specificity (SpPIn) rules in disease when positive. These are fundamental to evaluating population screening tests.

💡

Step 3 Pearl

When prevalence of a disease is very low, even a highly specific test will produce a large number of false positives, reducing the positive predictive value (PPV). This is why screening programs require careful selection of target populations to maximize PPV and clinical utility. Always consider prevalence when interpreting screening test results on the exam.

SECTION 5

Patient Safety & Quality Improvement Frameworks

Patient safety is a core dimension of population health that the USMLE Step 3 emphasizes heavily. The shift from a blame-oriented culture to a systems-based approach to error prevention has fundamentally changed how hospitals and clinics manage adverse events. Understanding the major safety models and quality improvement frameworks allows you to answer questions about root cause analysis, sentinel events, and continuous improvement cycles.

The Swiss Cheese Model shows that adverse events occur when latent failures (holes) in multiple defensive layers align. Each barrier—organizational culture, supervision, preconditions, specific safeguards, and point-of-care actions—has inherent weaknesses. Strengthening any single layer reduces the likelihood of error propagation through the entire system.

Quality Improvement Methodologies

Framework	Core Steps	Step 3 Application
Plan-Do-Study-Act (PDSA)	Plan a change → Implement on small scale → Study results → Act on findings (adopt, adapt, or abandon)	Rapid-cycle improvement in clinical settings; test new protocols iteratively
Root Cause Analysis (RCA)	Identify event → Gather data → Map contributing factors → Determine root cause → Implement corrective action	Post-sentinel-event investigation; required by Joint Commission for serious adverse events
Failure Mode and Effects Analysis (FMEA)	Prospectively identify potential failure modes → Rank by severity, probability, and detectability → Prioritize interventions	Proactive risk assessment; prevents errors before they occur
Lean / Six Sigma	Eliminate waste (Lean) and reduce process variation (Six Sigma) using DMAIC: Define → Measure → Analyze → Improve → Control	Hospital efficiency, reducing wait times, decreasing medication errors

Key Quality Improvement Frameworks Tested on USMLE Step 3

⚠

Sentinel Events

A sentinel event is an unexpected occurrence involving death or serious injury (or risk thereof) that is not related to the natural course of the patient's illness. Examples include wrong-site surgery, patient suicide in a monitored setting, and transfusion reactions from ABO-incompatible blood. The Joint Commission requires that institutions conduct a root cause analysis within 45 days of any sentinel event.

SECTION 6

Worked Example: Evaluating a Screening Program

A county health department considers implementing a new colorectal cancer screening program for adults aged 50–75. In a pilot study of 10,000 eligible adults, 200 are found to have colorectal cancer. The screening test detects 180 of these cases correctly and returns positive results for 400 people who do not have cancer. Calculate the sensitivity, specificity, positive predictive value, and determine the NNT if the screening program reduces colorectal cancer mortality from 5% to 3%.

Colorectal Cancer Screening Program Evaluation

1
Step 1 — Construct the 2×2 Table

From the data: TP = 180 (cancer detected), FN = 200 − 180 = 20 (cancer missed), FP = 400 (no cancer but positive test), TN = 9,800 − 400 = 9,400 (no cancer, negative test). Total population = 10,000; disease positive = 200; disease negative = 9,800.

TP = 180 | FN = 20 | FP = 400 | TN = 9,400
2
Step 2 — Calculate Sensitivity

Sensitivity = TP ÷ (TP + FN) = 180 ÷ (180 + 20) = 180 ÷ 200 = 0.90. The screening test correctly identifies 90% of individuals who truly have colorectal cancer.

Sensitivity = 90%
3
Step 3 — Calculate Specificity

Specificity = TN ÷ (TN + FP) = 9,400 ÷ (9,400 + 400) = 9,400 ÷ 9,800 ≈ 0.959. The test correctly identifies approximately 95.9% of individuals who do not have colorectal cancer.

Specificity ≈ 95.9%
4
Step 4 — Calculate Positive Predictive Value (PPV)

PPV = TP ÷ (TP + FP) = 180 ÷ (180 + 400) = 180 ÷ 580 ≈ 0.310. Only about 31% of patients who screen positive actually have colorectal cancer. This relatively low PPV reflects the low prevalence (2%) of disease in the screened population, demonstrating why prevalence critically affects PPV even when sensitivity and specificity are reasonably high.

PPV ≈ 31.0%
5
Step 5 — Calculate Number Needed to Screen (NNS)

If the screening program reduces colorectal cancer mortality from 5% to 3%, the absolute risk reduction (ARR) = 5% − 3% = 2% = 0.02. NNT (or NNS in screening contexts) = 1 ÷ ARR = 1 ÷ 0.02 = 50. This means you must screen 50 adults aged 50–75 for one additional life saved from colorectal cancer mortality.

NNS = 50 patients

🔍

CLINICAL INTERPRETATION

The low PPV (31%) does not mean the screening program is worthless—it means that many patients will undergo unnecessary follow-up colonoscopies. When deciding whether to implement the program, decision-makers must weigh the cost and risk of false positives (anxiety, procedural complications) against the benefit of detecting 90% of cancers early enough to reduce mortality. This cost-benefit calculus is the essence of population-level screening decisions.

SECTION 7

Strengths & Limitations of Population Health Approaches

No single approach to population health is without trade-offs. Understanding the strengths and limitations of different strategies helps physicians choose the most appropriate intervention and anticipate unintended consequences. The following comparison highlights the relative advantages and challenges of the major population health approaches tested on Step 3.

Approach	Strengths	Limitations
Universal Screening	Captures all affected individuals; reduces health disparities by not relying on risk-factor identification; standardizable protocols	High cost; increased false positives in low-prevalence populations; potential overdiagnosis and overtreatment
Targeted (High-Risk) Screening	Higher PPV; more cost-effective; reduced patient burden from unnecessary testing	May miss cases in perceived low-risk groups; risk-factor algorithms can perpetuate bias; requires accurate risk stratification tools
Policy/Environmental Interventions	Broadest population reach; do not require individual behavior change; cost-effective at scale (e.g., seatbelt laws, food fortification)	Political resistance; slow implementation; may not address individual variation; can be paternalistic
Clinical Preventive Services	Evidence-based (USPSTF grading); integrated into routine care; personalized to patient profile	Depends on healthcare access; provider adherence varies; limited by visit time and competing priorities
Health Education & Counseling	Empowers individual autonomy; supports shared decision-making; addresses health literacy	Least population impact (top of Frieden pyramid); behavior change is difficult to sustain; effectiveness varies by socioeconomic context

Population Health Approaches: Comparative Analysis

✦

KEY TAKEAWAY

Think of population health strategies as a diversified investment portfolio. Just as no single stock protects against all market risks, no single population health intervention addresses every health outcome. The most resilient health systems combine policy-level interventions (broad index funds), targeted screening (sector-specific investments), and individual clinical care (individual stock picks) to maximize overall population health returns while managing risks.

SECTION 8

Health Equity, Disparities, & Emerging Directions

Population health extends beyond traditional epidemiology into the domains of health equity and social justice. The USMLE Step 3 increasingly tests physicians on their understanding of health disparities—systematic differences in health outcomes that are avoidable, unjust, and linked to social disadvantage. Recognizing these disparities and understanding their root causes allows you to advocate for equitable care at both the individual and systems level.

Concept	Traditional View	Equity-Centered View
Access to Care	Insurance status and geographic proximity to providers	Structural barriers including transportation, language, implicit bias, historical mistrust, and culturally concordant care availability
Risk Factor Assessment	Individual behaviors (smoking, diet, exercise)	Upstream social determinants (food deserts, environmental exposures, adverse childhood experiences, systemic racism)
Quality Measurement	Aggregate population outcomes	Stratified outcomes by race, ethnicity, income, and geography to reveal hidden disparities within aggregate data
Intervention Design	One-size-fits-all programs	Targeted universalism: universal goals achieved through targeted strategies calibrated to each group's specific barriers

Evolution Toward Equity-Centered Population Health

Emerging areas that are increasingly relevant to Step 3 include the use of population health informatics (using electronic health record data and machine learning to identify at-risk populations), telehealth equity (ensuring digital health tools do not widen the digital divide), and climate change and health (recognizing how heat waves, air quality, and vector-borne disease patterns disproportionately affect vulnerable populations). These topics represent the frontier of population health science and will likely appear with increasing frequency on future iterations of the licensing examination.

⚠

Implicit Bias on Step 3

USMLE Step 3 may present clinical scenarios designed to test whether you apply different standards of care based on patient demographics. The correct answer is always to provide equitable, guideline-concordant care regardless of race, ethnicity, gender, or socioeconomic status. Be alert for answer choices that reflect stereotyping or differential treatment that is not supported by clinical evidence.

SECTION 9

Practice Problems

Question 1: CONCEPTUAL

A state health department is evaluating the burden of type 2 diabetes in its population. In 2023, there were 15,000 individuals living with type 2 diabetes in a population of 500,000. During that same year, 1,200 new cases of type 2 diabetes were diagnosed. A public health analyst reports that the point prevalence of type 2 diabetes in this population is 3%. Which of the following best explains why prevalence, rather than incidence, is the most appropriate measure to describe the 15,000 individuals currently living with type 2 diabetes?

A. Incidence rate
B. Prevalence captures all existing cases in the population at a point in time, regardless of when they were first diagnosed
C. Cumulative incidence
D. Mortality rate
E. Attack rate

Show answer

The correct answer is B. Prevalence refers to the total number of existing cases of a disease in a population at a given point in time (point prevalence) or over a period of time (period prevalence). The 15,000 individuals living with type 2 diabetes represent all existing cases — both old and new — which is the definition of prevalence. Prevalence = existing cases / total population = 15,000 / 500,000 = 0.03 or 3%, confirming the analyst's reported figure. Step-by-step verification: 1. Total existing cases = 15,000 2. Total population = 500,000 3. Point prevalence = 15,000 / 500,000 = 0.03 = 3% ✓ Incidence rate (A) measures the rate at which new cases develop in a population at risk over a specified time period; the 1,200 new diagnoses — not the 15,000 total cases — would be used to calculate incidence rate. Cumulative incidence (C) also refers only to new cases arising over a defined period in a disease-free population at the start; again, the 1,200 new diagnoses apply here, not the 15,000 total. Mortality rate (D) measures deaths occurring in a population over a time period and is unrelated to the number of people currently living with a disease. Attack rate (E) is a specialized form of cumulative incidence used in outbreak investigations to describe the proportion of exposed individuals who become ill; it is not applicable to chronic disease surveillance of existing cases.

Question 2: BASIC CALCULATION

A county health officer is investigating the effectiveness of a new community-based screening program for colorectal cancer. In the screened population of 10,000 adults aged 50-75, the screening test identified 200 individuals as positive. Of these 200, subsequent colonoscopy confirmed colorectal cancer in 50 individuals. Among the 9,800 individuals who screened negative, follow-up over the next year revealed 10 cases of colorectal cancer. What is the positive predictive value of this screening test?

A. 5%
B. 25%
C. 50%
D. 83%
E. 98%

Show answer

The correct answer is B. Positive predictive value (PPV) is calculated as the number of true positives divided by the total number of positive test results: PPV = TP / (TP + FP). In this scenario, 50 individuals were true positives (screened positive and confirmed cancer) and 150 were false positives (screened positive but no cancer on colonoscopy), so PPV = 50 / (50 + 150) = 50/200 = 0.25 or 25%. Answer A (5%) confuses PPV with the overall disease prevalence calculation. Answer C (50%) might result from dividing the true positives by the false positives rather than by all positives. Answer D (83%) represents the sensitivity of the test (50/60 = 83%), which is the proportion of all true disease cases correctly identified by the test. Answer E (98%) approximates the negative predictive value (9,790/9,800 = 99.9%) or specificity, which are different test characteristics from PPV.

Question 3: INTERMEDIATE

A 62-year-old woman presents to her primary care physician for an annual wellness visit. She has no significant past medical history and takes no medications. She has never had a colonoscopy. She does not smoke and drinks alcohol socially. Her family history is notable for a father who was diagnosed with colon cancer at age 72. Her BMI is 26 kg/m². She asks about cancer screening recommendations. According to current United States Preventive Services Task Force (USPSTF) guidelines, which of the following is the most appropriate screening recommendation for this patient?

A. No screening is needed since her father was diagnosed after age 70
B. Colonoscopy every 10 years or annual fecal immunochemical test (FIT)
C. Colonoscopy every 5 years due to her family history
D. CT colonography annually
E. Screening is no longer recommended after age 60

Show answer

The correct answer is B. The USPSTF recommends colorectal cancer screening for all adults aged 45-75 years (Grade A recommendation). Acceptable strategies include colonoscopy every 10 years, annual FIT, or other approved modalities. This patient is 62 and falls within the recommended screening age range. Answer A is incorrect because the USPSTF recommends average-risk screening for all adults aged 45-75 regardless of family history details; her father's diagnosis does not negate the need for her own screening. Answer C is incorrect because the USPSTF recommendations for average-risk individuals do not specify a 5-year colonoscopy interval based on a single first-degree relative diagnosed at age 72; specialty society guidelines (e.g., ACS or ACG) may recommend earlier or more frequent screening if a first-degree relative is diagnosed before age 60 or if multiple relatives are affected, but this scenario does not meet those criteria. Answer D is incorrect because CT colonography, while an accepted screening modality, is recommended every 5 years, not annually. Answer E is incorrect because the USPSTF recommends screening through age 75, with individualized decision-making for ages 76-85.

Question 4: APPLIED

A primary care physician in a rural community notices a cluster of 8 children aged 2-6 years presenting over 3 months with elevated blood lead levels (>5 µg/dL). All children attend the same daycare center, which is located in a building constructed in 1955. The physician reports the findings to the local health department. An environmental inspection reveals deteriorating lead-based paint on interior window sills and door frames of the daycare facility. Which of the following is the most appropriate next step in managing this public health concern?

A. Prescribe oral succimer (DMSA) chelation therapy for all affected children
B. Recommend the daycare center close temporarily and undergo professional lead abatement
C. Advise parents to increase dietary calcium intake in the children and recheck levels in 6 months
D. Recommend covering the deteriorating paint with a fresh coat of non-lead paint as a permanent solution
E. Perform hair and nail analysis on all children to confirm lead exposure before taking action

Show answer

The correct answer is B. When a clear environmental source of lead exposure is identified—in this case, deteriorating lead-based paint in a building constructed before 1978—the most critical public health intervention is source removal. Temporary closure of the daycare and professional lead abatement (removal or encapsulation by certified professionals) is necessary to prevent ongoing exposure to all children at the facility. Answer A is incorrect because chelation therapy is reserved for children with blood lead levels ≥45 µg/dL (or lower in symptomatic cases); it would not be appropriate for all children with levels >5 µg/dL, and it does not address the environmental source. Answer C is incorrect because while dietary calcium can reduce gastrointestinal lead absorption, it is an adjunctive measure and does not eliminate the ongoing exposure source; waiting 6 months would allow continued harmful exposure. Answer D is incorrect because simply painting over deteriorating lead paint is not an adequate or permanent solution, as the underlying paint will continue to deteriorate and can create lead dust, especially around friction surfaces like windows and doors. Answer E is incorrect because hair and nail analysis is not a reliable or recommended method for assessing lead exposure; blood lead levels are the standard diagnostic test, which has already been performed.

Question 5: CRITICAL THINKING

A city health department launches a mandatory reporting system for all cases of newly diagnosed hepatitis C. In the first year, 800 cases are reported among a city population of 400,000. The health department then implements a targeted screening program in high-risk populations (people who inject drugs, individuals born between 1945-1965, and incarcerated persons). In the second year, 1,400 cases of hepatitis C are reported. The city council expresses alarm at the apparent "increase" in hepatitis C and proposes diverting funds from the screening program to treatment infrastructure. A public health physician is asked to advise the council. Which of the following is the most accurate interpretation of the data to present to the city council?

A. The increase in reported cases represents a true increase in hepatitis C incidence requiring enhanced prevention efforts
B. The increase likely reflects improved case detection through targeted screening rather than a true rise in disease incidence
C. The screening program has failed because more cases were found, and resources should be redirected to treatment only
D. The data are unreliable because mandatory reporting systems consistently undercount cases and cannot detect real trends
E. The increase represents lead-time bias, and the screening program should be discontinued because it is detecting disease that would never cause symptoms

Show answer

The correct answer is B. When a new screening program is implemented, particularly one targeting high-risk populations, a predictable increase in detected (reported) cases occurs due to improved surveillance and case ascertainment—not necessarily a true increase in disease incidence. Many of these cases likely represent previously undiagnosed prevalent cases now being identified through active screening. This is a well-recognized phenomenon in public health surveillance. Answer A is incorrect because the temporal association between the new screening program and the rise in reported cases strongly suggests that enhanced detection, rather than a true increase in new infections, is the primary driver of the observed increase. Answer C is incorrect because finding more cases through screening is actually an indicator of program success, not failure; identifying hepatitis C cases allows linkage to treatment and prevention of transmission, and discontinuing screening would leave cases undiagnosed. Answer D is incorrect because while mandatory reporting systems may undercount cases, they can still detect meaningful trends, and the data are not inherently unreliable; moreover, the addition of active screening supplements passive reporting. Answer E is incorrect because hepatitis C is a progressive disease that causes significant morbidity (cirrhosis, hepatocellular carcinoma) if untreated; while lead-time bias is a valid concern in some screening contexts, it does not apply here since effective curative treatment (direct-acting antivirals) exists and early detection genuinely improves outcomes.

SUMMARY

Lesson Summary

Summary

Population health and safety represents the intersection of clinical medicine, epidemiology, health policy, and safety science. The Frieden health impact pyramid organizes interventions from broadest impact (socioeconomic changes and default modifications) to most individually intensive (clinical interventions and counseling). Key quantitative tools include incidence, prevalence, sensitivity and specificity, PPV and NPV, and number needed to treat (NNT)—all of which are high-yield for USMLE Step 3.

Patient safety depends on systems-based thinking as illustrated by the Swiss Cheese Model, supported by quality improvement frameworks such as PDSA cycles, root cause analysis, and FMEA. Finally, addressing health disparities and social determinants of health is essential for achieving equitable population health outcomes. Remember: on Step 3, always consider whether a question asks about an individual-level clinical decision or a population-level systems intervention, as the correct answer will differ accordingly.