Develop a research project using rhe case control method to study risk factors cardiovascular diseases in Pakistan. Describe how the study will be organised what data will be required and what methods of analysis will be used.

Reading File

Finding Sources

Searching PubMed

cardiovascular disease risk factors Pakistan epidemiology

Searching the Web

case control study cardiovascular disease risk factors Pakistan research methodology

Finding Sources

Reading Web Page

https://pmc.ncbi.nlm.nih.gov/articles/PMC6284873

Reading File

Loading Image

I now have everything needed to write a thorough, well-structured research project proposal. Here it is:

Case-Control Study: Risk Factors for Cardiovascular Diseases in Pakistan

Research Project Proposal

1. Background and Rationale

Cardiovascular diseases (CVDs) are the leading cause of mortality in Pakistan. Epidemiological surveys from tertiary hospitals in Islamabad, Karachi, and Nawabshah have consistently documented a high burden of modifiable risk factors - hypertension (found in 66-73% of CVD patients), dyslipidaemia (64%), smoking (66%), obesity (52%), and diabetes (31%) - within CVD case populations. Pakistan's demographic profile (rapid urbanisation, dietary transitions, physical inactivity, and high rates of tobacco use) makes the identification and quantification of locally relevant risk factors a public health priority.

A case-control design is appropriate here because CVD is a chronic disease where the causal pathway may span decades, the disease has already developed in cases, and the method allows study of multiple exposures simultaneously at low cost - the standard first approach to testing causal hypotheses for serious conditions, as described by Park's Textbook of Preventive and Social Medicine.

2. Study Design

Design: Hospital-based, multi-centre case-control study Direction of Inquiry: Retrospective (from effect to cause)

The study proceeds backwards in time: both exposure and outcome have already occurred before the study begins. The investigator starts with persons who have CVD (cases) and persons who do not (controls), then looks back at their histories for prior exposure to suspected risk factors.

Case-Control Study Design - Direction of inquiry is backwards from effect (cases/controls) to cause (exposure status)

Fig. 1: Schematic design of a case-control study - Park's Textbook of Preventive and Social Medicine

Case-to-control ratio: 1:1 (or up to 1:2 if controls are readily available, to increase statistical power without proportional cost increase)

3. Study Objectives

Primary objective: Identify and quantify the association between established risk factors and the occurrence of CVD (coronary artery disease and acute myocardial infarction) in the Pakistani population.

Secondary objectives:

Estimate the odds ratio (OR) for each risk factor
Determine the population attributable risk (PAR) for modifiable risk factors
Examine the interaction between multiple co-existing risk factors
Compare risk factor profiles across sex, urban vs. rural residence, and socioeconomic strata

4. Study Organisation

4.1 Study Sites

A multi-centre approach spanning three major cities to capture geographic and socioeconomic diversity:

Lahore: Services Hospital / Punjab Institute of Cardiology (high-volume tertiary cardiac centre)
Karachi: Aga Khan University Hospital / National Institute of Cardiovascular Diseases (NICVD)
Islamabad/Rawalpindi: Pakistan Institute of Medical Sciences (PIMS) / Shifa International Hospital

This mirrors the approach of Liaquat et al. (2018), who recruited from PIMS and Shifa International, finding significantly deranged lipid profiles, elevated BMI, and hypertension in the case group vs. controls (p < 0.05).

4.2 Sample Size

Using a standard power calculation (80% power, alpha = 0.05, two-tailed), based on an estimated OR of 2.0 for hypertension with 30% exposure in controls:

Minimum: ~400 cases + 400 controls per centre
Total target: ~800-1000 cases, 800-1000 controls across all centres

This is consistent with published Pakistani case-control CVD studies (Liaquat et al.: 835 cases, 794 controls).

4.3 Study Period

18 months total:

3 months: ethics approval, protocol finalisation, training of data collectors, pilot testing
12 months: data collection (recruitment of cases and controls)
3 months: data cleaning, analysis, and write-up

5. Selection of Cases and Controls

5.1 Case Definition

Following Park's principle that a prior definition of what constitutes a "case" is fundamental, cases must satisfy both:

(a) Diagnostic criteria - confirmed incident CVD defined as any of:

Acute myocardial infarction: ischaemic symptoms + ECG changes (ST elevation/depression, new LBBB) + elevated cardiac troponins (troponin I or T above the 99th percentile)
Unstable angina / NSTEMI: clinical diagnosis with ECG and enzyme criteria
Coronary artery disease confirmed on coronary angiography (≥50% stenosis in at least one major vessel)

(b) Eligibility criteria:

Adults aged 30-70 years, newly diagnosed (incident cases, not prevalent)
Admitted to the cardiology department or CCU of study hospitals
Willing to provide informed consent
Resident of Pakistan for at least 5 years

Exclusion criteria for cases: Congenital heart disease, valvular disease as primary diagnosis, rheumatic heart disease, cardiomyopathy unrelated to ischaemia.

5.2 Control Selection

Controls are the most challenging component. Per Park's guidelines, controls must be:

Free of the disease under study (no history of CVD, confirmed by clinical assessment and ECG)
Comparable to cases on key confounding variables (age ± 5 years, sex, hospital/locality)
Representative of the source population from which cases arose

Sources of controls:

General outpatient department (OPD) attendees from the same hospitals attending for unrelated conditions (orthopaedic, dermatology, ENT)
Community volunteers and medical staff from the same catchment area

One control per case will be individually matched on age (±5 years) and sex to control for these strong confounders.

6. Matching

Individual (pair) matching on:

Age (±5 years)
Sex
Hospital (to control for facility-level referral bias)

Matching on socioeconomic status and geographic area (urban/rural) will be done at group level through stratified analysis rather than individual matching, to preserve statistical flexibility and avoid over-matching.

As Park notes, cases and controls must be comparable with respect to known confounding factors such as age, sex, occupation, and social status.

7. Data Collection - Exposures and Variables

7.1 Data Required

Data will be collected through a structured interviewer-administered questionnaire, clinical examination, and laboratory investigations.

A. Sociodemographic variables

Age, sex, education level, occupation
Monthly household income (socioeconomic class)
Urban vs. rural residence
Ethnicity/province of origin (Punjabi, Sindhi, Pashtun, Baloch, Muhajir)

B. Clinical measurements (at enrolment)

Measurement	Method/Criterion
Blood pressure	Two readings, 5 min apart; hypertension = SBP ≥140 or DBP ≥90 mmHg, or on antihypertensives
Body weight, height	Calibrated scales; BMI = weight(kg)/height(m)²
Waist circumference	Measured at umbilicus; central obesity: >90 cm men, >80 cm women (South Asian cut-offs)
Fasting blood glucose	Capillary or venous; diabetes = FBG ≥7.0 mmol/L or on hypoglycaemics

C. Biochemical investigations (fasting, 8-12 hours)

Total cholesterol, LDL-cholesterol, HDL-cholesterol, triglycerides
Fasting blood glucose, HbA1c
Serum creatinine (renal function)
Full blood count (anaemia as confounder)
High-sensitivity C-reactive protein (hs-CRP) as inflammatory marker

D. Lifestyle and behavioural risk factors (questionnaire)

Smoking: current, ex-smoker (quit >1 year), never; pack-year history; smokeless tobacco (naswar/gutka use - common in Pakistan)
Physical activity: measured by IPAQ (International Physical Activity Questionnaire) - sedentary defined as <600 MET-minutes/week
Dietary history: frequency intake of red meat, fried food, vegetables, fruits, salt added to food, ghee/oil use
Alcohol consumption (culturally sensitive; low prevalence in Pakistan but must be assessed)
Sleep patterns and psychosocial stress (perceived stress scale)

E. Past medical and family history

Duration of hypertension and diabetes
Family history of premature CVD (first-degree relative with MI or stroke before age 55 in males, 65 in females)
Previous TIA or stroke
History of chronic kidney disease

F. Medication history

Current antihypertensive therapy, statins, antidiabetic agents, antiplatelet drugs

7.2 Data Collection Tools

Pre-tested, bilingual (English/Urdu) structured questionnaire
Calibrated digital sphygmomanometers (validated for Pakistani populations)
Standard 12-lead ECG for all controls (to exclude silent ischaemia)
Venous blood sample collection by trained phlebotomists; samples processed in a single centralised laboratory per centre using standardised methods

8. Ethical Considerations

Ethics approval from Institutional Review Boards (IRB) of all participating hospitals and from the Pakistan Medical Research Council (PMRC)
Written informed consent from all participants (Urdu-language consent forms)
Anonymisation: participants identified by study ID codes only
Right to withdraw at any time without consequence
All laboratory results shared with participants and treating physicians
Data stored on password-protected servers; hard copies in locked cabinets

9. Methods of Analysis

9.1 The 2×2 Contingency Table

The fundamental analytical framework is the 2×2 table, applied to each risk factor separately:

	Cases (CVD)	Controls (No CVD)
Risk factor present	a	b
Risk factor absent	c	d
Total	a+c	b+d

Exposure rates: Cases = a/(a+c); Controls = b/(b+d)

(Park's Textbook of Preventive and Social Medicine, p. 83-84)

9.2 Odds Ratio (OR) - Primary Measure of Association

Because incidence rates cannot be directly calculated from a case-control study (since the investigator chooses the number of cases and controls), the Odds Ratio is used as the measure of association:

OR = (a × d) / (b × c)

OR = 1.0: no association
OR > 1.0: positive association (risk factor)
OR < 1.0: protective factor

The OR approximates the Relative Risk (RR) when the disease is rare in the population - a valid assumption for incident CVD in a given year.

95% Confidence intervals will be calculated for each OR using the Woolf method or logistic regression.

9.3 Tests of Statistical Significance

Chi-square (χ²) test for categorical variables (e.g., smoking yes/no, hypertension yes/no)
Student's independent t-test for continuous variables (e.g., mean total cholesterol, mean systolic blood pressure) - as used by Liaquat et al. in Islamabad
Mann-Whitney U test if continuous variables are not normally distributed

Significance threshold: p < 0.05 (two-tailed). As Park notes, statistical association does not imply causation; p-values are a starting point, not the endpoint of analysis.

9.4 Matched Analysis

Since individual 1:1 matching is used:

McNemar's test for dichotomous matched pairs
Conditional logistic regression for matched data to calculate adjusted ORs

9.5 Multivariable Analysis - Logistic Regression

Unconditional (or conditional) logistic regression is the primary multivariable method. This allows:

Simultaneous adjustment for multiple confounders
Calculation of adjusted ORs for each risk factor, holding all others constant
Assessment of interaction terms (e.g., does the effect of smoking differ between diabetics and non-diabetics?)

Variables entered into the model:

All risk factors significant at p < 0.20 in univariate analysis
A priori confounders: age, sex, socioeconomic status, family history

9.6 Population Attributable Risk (PAR)

PAR quantifies how much of the disease burden in Pakistan's population is attributable to a specific factor:

PAR% = [Pe(OR - 1)] / [1 + Pe(OR - 1)] × 100

Where Pe = prevalence of the exposure in the population.

This is the most policy-relevant statistic - it identifies which risk factor, if eliminated, would achieve the greatest reduction in CVD burden nationally.

9.7 Stratified Analysis (Mantel-Haenszel)

Stratified analysis using the Mantel-Haenszel method will be used to:

Examine effect modification by sex, age group, urban/rural status
Detect confounding (comparing crude ORs to stratum-specific ORs)

9.8 Sensitivity Analyses

Separate analyses restricting to incident (new) cases only
Analyses excluding participants with missing data
Analyses restricting to age 30-60 (premature CVD, particularly relevant in South Asia)

10. Bias and Its Control

Bias Type	Source	Control
Selection bias	Hospital controls may not represent the general population	Use multiple control sources; community controls as sensitivity analysis
Recall bias	Cases may over-report past exposures (since they know their diagnosis)	Use objective measures (blood tests, medical records) not only self-report
Observer bias	Interviewers may probe differently for cases vs. controls	Interviewers blinded to case/control status where possible; standardised questionnaire
Information bias	Inaccurate dietary recall	Validated food frequency questionnaires in Urdu
Confounding	Socioeconomic status confounds many exposures	Matching + logistic regression adjustment

11. Expected Risk Factors and Hypotheses

Based on prior Pakistani and South Asian data, the study hypothesises that the following factors will show significantly elevated ORs:

Risk Factor	Expected Direction	Biological Basis
Hypertension	OR > 3	Endothelial injury, LVH, atherosclerosis
Dyslipidaemia (high LDL, low HDL)	OR > 2.5	Atherogenesis via foam cell formation
Current smoking / tobacco use	OR > 2	Oxidative stress, platelet activation, vasospasm
Type 2 diabetes	OR > 2	Accelerated atherosclerosis, pro-thrombotic state
Abdominal obesity	OR > 1.8	Insulin resistance, pro-inflammatory adipokines
Physical inactivity	OR > 1.5	Reduced HDL, impaired endothelial function
Family history of premature CVD	OR > 2	Polygenic susceptibility
High dietary saturated fat/ghee	OR > 1.5	Elevated LDL cholesterol
Psychosocial stress	OR > 1.4	Sympathoadrenal activation, cortisol
Smokeless tobacco (naswar)	OR > 1.6	Nicotine-mediated vasoconstriction - unique to South Asia

12. Strengths and Limitations of the Design

Strengths:

Faster and cheaper than a cohort study - no need to wait for disease to develop
Suitable for rare outcomes (CVD events)
Can study multiple risk factors simultaneously
Multi-centre design improves external validity across Pakistan's diverse regions
Inclusion of smokeless tobacco (a Pakistan-specific exposure often missed in Western studies)

Limitations:

Cannot calculate incidence rates or absolute risk
Retrospective exposure assessment is subject to recall bias
Hospital-based controls may share some risk factors with cases (Berkson's bias)
Cannot establish temporality with certainty (though for risk factors like hypertension, pre-existing records help)
Case-control studies yield OR, not RR directly (though OR approximates RR when disease prevalence is low)

13. Dissemination and Policy Impact

Findings will be submitted to peer-reviewed journals (JPMA, Eastern Mediterranean Health Journal) and presented to the Pakistan Ministry of National Health Services. PAR estimates will directly inform the prioritisation of public health interventions - for example, if hypertension accounts for 35% of attributable risk, this strengthens the case for community-level blood pressure screening programmes.

Summary Table

Component	Detail
Design	Hospital-based case-control
Outcome	Incident CVD (MI, CAD, unstable angina)
Sites	Lahore, Karachi, Islamabad
Sample size	~900 cases + 900 controls
Matching	Individual: age (±5 yr), sex; Group: hospital
Exposure data	Questionnaire + clinical exam + fasting bloods
Primary statistic	Odds Ratio (with 95% CI)
Univariate analysis	χ² test, t-test
Multivariable analysis	Conditional logistic regression
Policy measure	Population Attributable Risk %

Key references:

Park's Textbook of Preventive and Social Medicine outlines the four basic steps of case-control design (case/control selection, matching, exposure measurement, analysis) and the OR/PAR framework used throughout this protocol.
Liaquat et al. (2018) carried out a case-control study in Islamabad (835 cases, 794 controls) finding significantly elevated BMI, SBP, DBP, and dyslipidaemia in CVD patients vs. controls (p < 0.05) - providing baseline OR estimates for power calculations.
Jamali et al. (2024) (Nawabshah case-control study) documented hypertension (73.5%), smoking (66.4%), dyslipidaemia (64.3%), and obesity (52.5%) as the dominant modifiable risk factors in Pakistani CVD patients.

Generate a power point presentation of the above research and mention on first slide Name: SHAHAB Group: Med21C

Reading File

Running Command

mkdir -p /tmp/workspace/cvd-pakistan-pptx

Writing File

~/cvd-pakistan-pptx/create_pptx.py

from pptx import Presentation
from pptx.util import Inches, Pt, Emu
from pptx.dml.color import RGBColor
from pptx.enum.text import PP_ALIGN, MSO_ANCHOR
from pptx.util import Inches, Pt
import copy

# ── Colour palette ──────────────────────────────────────────────────────────
DARK_RED   = RGBColor(0xC0, 0x00, 0x00)
DEEP_NAVY  = RGBColor(0x1F, 0x34, 0x64)
MID_BLUE   = RGBColor(0x2E, 0x75, 0xB6)
LIGHT_BLUE = RGBColor(0xBD, 0xD7, 0xEE)
WHITE      = RGBColor(0xFF, 0xFF, 0xFF)
LIGHT_GREY = RGBColor(0xF2, 0xF2, 0xF2)
DARK_GREY  = RGBColor(0x40, 0x40, 0x40)
ACCENT_RED = RGBColor(0xC0, 0x00, 0x00)

prs = Presentation()
prs.slide_width  = Inches(13.333)
prs.slide_height = Inches(7.5)

blank = prs.slide_layouts[6]  # fully blank


# ── Helper functions ─────────────────────────────────────────────────────────
def add_rect(slide, x, y, w, h, fill_color, alpha=None):
    shape = slide.shapes.add_shape(
        1,  # MSO_SHAPE_TYPE.RECTANGLE
        Inches(x), Inches(y), Inches(w), Inches(h)
    )
    shape.line.fill.background()
    fill = shape.fill
    fill.solid()
    fill.fore_color.rgb = fill_color
    return shape


def add_text(slide, text, x, y, w, h, size, bold=False, color=WHITE,
             align=PP_ALIGN.LEFT, italic=False, wrap=True, font="Calibri"):
    tb = slide.shapes.add_textbox(Inches(x), Inches(y), Inches(w), Inches(h))
    tf = tb.text_frame
    tf.word_wrap = wrap
    tf.margin_left = Pt(0)
    tf.margin_right = Pt(0)
    tf.margin_top = Pt(0)
    tf.margin_bottom = Pt(0)
    p = tf.paragraphs[0]
    p.alignment = align
    run = p.add_run()
    run.text = text
    run.font.name = font
    run.font.size = Pt(size)
    run.font.bold = bold
    run.font.italic = italic
    run.font.color.rgb = color
    return tb


def add_bullet_slide(slide, title_text, bullets, bg_color=LIGHT_GREY):
    """Standard content slide with header bar + bullet list."""
    # Background
    add_rect(slide, 0, 0, 13.333, 7.5, bg_color)
    # Header bar
    add_rect(slide, 0, 0, 13.333, 1.15, DEEP_NAVY)
    # Accent strip
    add_rect(slide, 0, 1.15, 13.333, 0.07, DARK_RED)
    # Title
    add_text(slide, title_text, 0.35, 0.18, 12.5, 0.85,
             size=28, bold=True, color=WHITE, align=PP_ALIGN.LEFT)
    # Bullets
    tb = slide.shapes.add_textbox(Inches(0.5), Inches(1.4), Inches(12.3), Inches(5.8))
    tf = tb.text_frame
    tf.word_wrap = True
    tf.margin_left = Pt(4)
    for i, (lvl, txt) in enumerate(bullets):
        p = tf.paragraphs[0] if i == 0 else tf.add_paragraph()
        p.level = lvl
        run = p.add_run()
        run.text = txt
        run.font.name = "Calibri"
        run.font.size = Pt(17) if lvl == 0 else Pt(15)
        run.font.bold = (lvl == 0)
        run.font.color.rgb = DEEP_NAVY if lvl == 0 else DARK_GREY
        p.space_before = Pt(5 if lvl == 0 else 2)
        p.space_after  = Pt(2)


# ═══════════════════════════════════════════════════════════════════════════════
# SLIDE 1 — Title Slide
# ═══════════════════════════════════════════════════════════════════════════════
slide = prs.slides.add_slide(blank)

# Full dark navy background
add_rect(slide, 0, 0, 13.333, 7.5, DEEP_NAVY)
# Top red accent stripe
add_rect(slide, 0, 0, 13.333, 0.18, DARK_RED)
# Bottom red accent stripe
add_rect(slide, 0, 7.22, 13.333, 0.28, DARK_RED)
# Mid decorative bar
add_rect(slide, 0, 3.8, 13.333, 0.06, MID_BLUE)
# White vertical side accent
add_rect(slide, 0, 0.18, 0.12, 7.04, MID_BLUE)

# Main title
add_text(slide,
         "Risk Factors for Cardiovascular Diseases in Pakistan",
         0.5, 0.55, 12.3, 2.2,
         size=38, bold=True, color=WHITE, align=PP_ALIGN.CENTER)

# Subtitle
add_text(slide,
         "A Case-Control Study Design",
         0.5, 2.85, 12.3, 0.8,
         size=26, bold=False, color=LIGHT_BLUE, align=PP_ALIGN.CENTER, italic=True)

# Divider line (thin rectangle)
add_rect(slide, 2.5, 3.9, 8.333, 0.04, LIGHT_BLUE)

# Student info box
add_rect(slide, 3.0, 4.1, 7.2, 2.1, MID_BLUE)
add_text(slide, "Presented by:",       3.1, 4.18, 7.0, 0.45, size=16, color=LIGHT_BLUE, align=PP_ALIGN.CENTER, italic=True)
add_text(slide, "SHAHAB",              3.1, 4.6,  7.0, 0.65, size=30, bold=True,  color=WHITE, align=PP_ALIGN.CENTER)
add_text(slide, "Group: Med21C",       3.1, 5.25, 7.0, 0.45, size=20, color=LIGHT_BLUE, align=PP_ALIGN.CENTER)
add_text(slide, "July 2026",           3.1, 5.75, 7.0, 0.35, size=15, color=LIGHT_BLUE, align=PP_ALIGN.CENTER, italic=True)


# ═══════════════════════════════════════════════════════════════════════════════
# SLIDE 2 — Background & Rationale
# ═══════════════════════════════════════════════════════════════════════════════
slide = prs.slides.add_slide(blank)
add_bullet_slide(slide, "Background & Rationale", [
    (0, "CVDs: The Leading Killer in Pakistan"),
    (1, "Cardiovascular diseases are the #1 cause of mortality in Pakistan"),
    (1, "Rapid urbanisation, dietary transitions & tobacco use are accelerating the burden"),
    (1, "Pakistan has a unique epidemiological profile distinct from Western populations"),
    (0, "Why a Case-Control Study?"),
    (1, "Efficient & cost-effective for studying rare outcomes like acute CVD events"),
    (1, "Can examine multiple risk factors simultaneously"),
    (1, "Ideal for chronic diseases where causal pathway spans decades"),
    (1, "Both exposure and outcome have already occurred — study works backwards"),
])


# ═══════════════════════════════════════════════════════════════════════════════
# SLIDE 3 — Study Objectives
# ═══════════════════════════════════════════════════════════════════════════════
slide = prs.slides.add_slide(blank)
add_bullet_slide(slide, "Study Objectives", [
    (0, "Primary Objective"),
    (1, "Identify and quantify associations between established risk factors and CVD"),
    (1, "Calculate Odds Ratios (ORs) for each exposure with 95% confidence intervals"),
    (0, "Secondary Objectives"),
    (1, "Determine Population Attributable Risk (PAR%) for modifiable risk factors"),
    (1, "Examine interactions between co-existing risk factors"),
    (1, "Compare risk factor profiles by sex, age group, urban vs. rural residence"),
    (1, "Identify Pakistan-specific factors (e.g. smokeless tobacco — naswar/gutka)"),
])


# ═══════════════════════════════════════════════════════════════════════════════
# SLIDE 4 — Study Design Overview
# ═══════════════════════════════════════════════════════════════════════════════
slide = prs.slides.add_slide(blank)
add_bullet_slide(slide, "Study Design Overview", [
    (0, "Design:   Hospital-based, multi-centre Case-Control Study"),
    (0, "Study Sites"),
    (1, "Lahore — Punjab Institute of Cardiology / Services Hospital"),
    (1, "Karachi — NICVD / Aga Khan University Hospital"),
    (1, "Islamabad — PIMS / Shifa International Hospital"),
    (0, "Sample Size"),
    (1, "~900 Cases  +  ~900 Controls  (1:1 ratio)"),
    (1, "Power: 80%  |  Alpha: 0.05  |  Expected OR: 2.0"),
    (0, "Study Duration:  18 months total"),
    (1, "3 months — ethics, protocol, training  |  12 months — recruitment  |  3 months — analysis"),
])


# ═══════════════════════════════════════════════════════════════════════════════
# SLIDE 5 — Case-Control Design Diagram
# ═══════════════════════════════════════════════════════════════════════════════
slide = prs.slides.add_slide(blank)
add_rect(slide, 0, 0, 13.333, 7.5, LIGHT_GREY)
add_rect(slide, 0, 0, 13.333, 1.15, DEEP_NAVY)
add_rect(slide, 0, 1.15, 13.333, 0.07, DARK_RED)
add_text(slide, "Case-Control Study: Conceptual Framework",
         0.35, 0.18, 12.5, 0.85, size=28, bold=True, color=WHITE)

# Direction labels
add_text(slide, "TIME →", 4.5, 1.5, 4, 0.4, size=14, color=DARK_GREY, bold=True, align=PP_ALIGN.CENTER)
add_text(slide, "← Direction of Inquiry", 4.5, 1.9, 4, 0.4, size=13, color=DARK_RED, bold=True, align=PP_ALIGN.CENTER, italic=True)

# POPULATION box
add_rect(slide, 5.5, 2.6, 2.3, 0.85, DEEP_NAVY)
add_text(slide, "POPULATION", 5.5, 2.7, 2.3, 0.65, size=14, bold=True, color=WHITE, align=PP_ALIGN.CENTER)

# CASES box
add_rect(slide, 2.2, 1.95, 2.5, 0.85, DARK_RED)
add_text(slide, "CASES\n(CVD present)", 2.2, 1.98, 2.5, 0.8, size=13, bold=True, color=WHITE, align=PP_ALIGN.CENTER)

# CONTROLS box
add_rect(slide, 8.6, 1.95, 2.5, 0.85, MID_BLUE)
add_text(slide, "CONTROLS\n(CVD absent)", 8.6, 1.98, 2.5, 0.8, size=13, bold=True, color=WHITE, align=PP_ALIGN.CENTER)

# EXPOSED / NOT EXPOSED for cases
add_rect(slide, 0.3, 1.6, 1.6, 0.6, ACCENT_RED)
add_text(slide, "Exposed", 0.3, 1.68, 1.6, 0.45, size=13, bold=True, color=WHITE, align=PP_ALIGN.CENTER)
add_rect(slide, 0.3, 2.45, 1.6, 0.6, RGBColor(0xE0,0x70,0x70))
add_text(slide, "Not Exposed", 0.3, 2.53, 1.6, 0.45, size=12, bold=False, color=WHITE, align=PP_ALIGN.CENTER)

# EXPOSED / NOT EXPOSED for controls
add_rect(slide, 11.4, 1.6, 1.6, 0.6, MID_BLUE)
add_text(slide, "Exposed", 11.4, 1.68, 1.6, 0.45, size=13, bold=True, color=WHITE, align=PP_ALIGN.CENTER)
add_rect(slide, 11.4, 2.45, 1.6, 0.6, RGBColor(0x90,0xB8,0xD8))
add_text(slide, "Not Exposed", 11.4, 2.53, 1.6, 0.45, size=12, bold=False, color=WHITE, align=PP_ALIGN.CENTER)

# Key principle box
add_rect(slide, 0.5, 3.7, 12.3, 1.35, DEEP_NAVY)
add_text(slide,
         "Key Principle: Start with DISEASE STATUS (cases vs. controls), then look BACK in time to compare exposure histories",
         0.6, 3.78, 12.1, 1.2, size=16, bold=False, color=WHITE, align=PP_ALIGN.CENTER)

# 2x2 table explanation
add_rect(slide, 0.5, 5.2, 12.3, 2.0, WHITE)
add_text(slide, "The 2×2 Contingency Table:", 0.65, 5.25, 5, 0.4, size=14, bold=True, color=DEEP_NAVY)
# table headers
cols = ["", "Cases (CVD +)", "Controls (CVD −)"]
xs = [0.65, 3.8, 7.8]
for j, (x, h) in enumerate(zip(xs, cols)):
    add_text(slide, h, x, 5.65, 3.5, 0.4, size=13, bold=True, color=DEEP_NAVY)
rows = [("Risk Factor Present", "a", "b"), ("Risk Factor Absent", "c", "d")]
ys = [6.05, 6.5]
for row, y in zip(rows, ys):
    for x, cell in zip(xs, row):
        add_text(slide, cell, x, y, 3.5, 0.4, size=13, color=DARK_GREY)
add_text(slide, "  Odds Ratio  =  (a × d) / (b × c)", 7.5, 5.65, 5.0, 1.3, size=15, bold=True, color=DARK_RED)


# ═══════════════════════════════════════════════════════════════════════════════
# SLIDE 6 — Case & Control Selection
# ═══════════════════════════════════════════════════════════════════════════════
slide = prs.slides.add_slide(blank)
add_bullet_slide(slide, "Selection of Cases & Controls", [
    (0, "Cases — Inclusion Criteria"),
    (1, "Adults aged 30–70 years with NEWLY diagnosed CVD (incident cases)"),
    (1, "Diagnosis confirmed by: ECG changes + elevated troponins + clinical assessment"),
    (1, "OR: CAD confirmed on coronary angiography (≥50% stenosis)"),
    (1, "Admitted to cardiology/CCU of study hospitals | Resident of Pakistan ≥5 years"),
    (0, "Controls — Selection Criteria"),
    (1, "No history or clinical evidence of CVD (normal ECG at enrolment)"),
    (1, "Recruited from General OPD (orthopaedics, dermatology, ENT)"),
    (1, "Individually matched: age (±5 years) + sex + hospital site"),
    (0, "Matching Variables"),
    (1, "Individual: Age (±5 yr), Sex  |  Group: Socioeconomic strata, Urban/Rural"),
])


# ═══════════════════════════════════════════════════════════════════════════════
# SLIDE 7 — Data Required
# ═══════════════════════════════════════════════════════════════════════════════
slide = prs.slides.add_slide(blank)
add_bullet_slide(slide, "Data Required — Variables & Measurements", [
    (0, "A. Sociodemographic"),
    (1, "Age, sex, education, occupation, income level, urban/rural, ethnicity/province"),
    (0, "B. Clinical Measurements"),
    (1, "Blood pressure (hypertension = SBP ≥140 / DBP ≥90 mmHg or on treatment)"),
    (1, "BMI (kg/m²) | Waist circumference (central obesity: >90 cm M, >80 cm F)"),
    (0, "C. Biochemical (Fasting 8–12 hrs)"),
    (1, "Lipid profile: Total-C, LDL-C, HDL-C, Triglycerides | FBG & HbA1c | hs-CRP"),
    (0, "D. Lifestyle Factors (Questionnaire)"),
    (1, "Smoking (pack-years) | Smokeless tobacco (naswar/gutka) | Physical activity (IPAQ)"),
    (1, "Dietary intake (ghee, red meat, salt, vegetables) | Psychosocial stress scale"),
    (0, "E. Medical & Family History"),
    (1, "Duration of hypertension/diabetes | Family history of premature CVD | Medications"),
])


# ═══════════════════════════════════════════════════════════════════════════════
# SLIDE 8 — Expected Risk Factors & Hypotheses
# ═══════════════════════════════════════════════════════════════════════════════
slide = prs.slides.add_slide(blank)
add_rect(slide, 0, 0, 13.333, 7.5, LIGHT_GREY)
add_rect(slide, 0, 0, 13.333, 1.15, DEEP_NAVY)
add_rect(slide, 0, 1.15, 13.333, 0.07, DARK_RED)
add_text(slide, "Hypothesised Risk Factors & Expected Odds Ratios",
         0.35, 0.18, 12.5, 0.85, size=26, bold=True, color=WHITE)

# Table
headers = ["Risk Factor", "Expected OR", "Biological Basis"]
col_xs  = [0.35, 5.2, 7.5]
col_ws  = [4.7,  2.1, 5.5]
row_data = [
    ("Hypertension",            "> 3.0",  "Endothelial injury → atherosclerosis"),
    ("Dyslipidaemia (high LDL)", "> 2.5", "Foam cell formation → plaque"),
    ("Smoking / Tobacco use",   "> 2.0",  "Oxidative stress, platelet activation"),
    ("Type 2 Diabetes",         "> 2.0",  "Accelerated atherosclerosis"),
    ("Abdominal Obesity",       "> 1.8",  "Insulin resistance, pro-inflammatory"),
    ("Physical Inactivity",     "> 1.5",  "Reduced HDL, endothelial dysfunction"),
    ("Family History CVD",      "> 2.0",  "Polygenic susceptibility"),
    ("Naswar / Smokeless Tobacco", "> 1.6", "Nicotine → vasoconstriction (South Asia-specific)"),
    ("Psychosocial Stress",     "> 1.4",  "Sympathoadrenal activation"),
]

# Header row
y = 1.35
add_rect(slide, 0.3, y, 12.7, 0.42, DEEP_NAVY)
for x, w, h in zip(col_xs, col_ws, headers):
    add_text(slide, h, x, y+0.04, w, 0.36, size=13, bold=True, color=WHITE)

for i, row in enumerate(row_data):
    y += 0.5
    bg = WHITE if i % 2 == 0 else LIGHT_BLUE
    add_rect(slide, 0.3, y, 12.7, 0.46, bg)
    for x, w, cell in zip(col_xs, col_ws, row):
        is_or = (cell.startswith(">"))
        add_text(slide, cell, x, y+0.04, w, 0.4, size=12,
                 bold=is_or, color=ACCENT_RED if is_or else DARK_GREY)


# ═══════════════════════════════════════════════════════════════════════════════
# SLIDE 9 — Methods of Analysis
# ═══════════════════════════════════════════════════════════════════════════════
slide = prs.slides.add_slide(blank)
add_bullet_slide(slide, "Methods of Analysis", [
    (0, "1. Descriptive Statistics"),
    (1, "Means ± SD for continuous variables | Frequencies (%) for categorical variables"),
    (0, "2. Univariate Analysis"),
    (1, "Chi-square (χ²) test — categorical exposures (smoking, hypertension, diabetes)"),
    (1, "Student's independent t-test — continuous variables (BP, cholesterol, BMI)"),
    (1, "Crude Odds Ratio = (a × d) / (b × c) with 95% Confidence Intervals"),
    (0, "3. Matched Analysis (for paired data)"),
    (1, "McNemar's test | Conditional logistic regression (adjusts for matching)"),
    (0, "4. Multivariable Logistic Regression"),
    (1, "Adjusted ORs controlling for all confounders simultaneously"),
    (1, "Interaction terms to detect effect modification (e.g. smoking × diabetes)"),
    (0, "5. Population Attributable Risk (PAR%)"),
    (1, "PAR% = [Pe(OR-1)] / [1+Pe(OR-1)] × 100 — identifies highest-impact targets"),
])


# ═══════════════════════════════════════════════════════════════════════════════
# SLIDE 10 — Bias Control & Ethical Considerations
# ═══════════════════════════════════════════════════════════════════════════════
slide = prs.slides.add_slide(blank)
add_bullet_slide(slide, "Bias Control & Ethical Considerations", [
    (0, "Controlling Major Biases"),
    (1, "Selection bias → Multiple control sources; community controls as sensitivity analysis"),
    (1, "Recall bias → Objective measures (lab tests, medical records) not just self-report"),
    (1, "Observer bias → Interviewers blinded to case/control status; standardised tool"),
    (1, "Confounding → Individual matching + multivariable logistic regression"),
    (0, "Ethical Safeguards"),
    (1, "Ethics approval: IRBs of all 3 centres + Pakistan Medical Research Council (PMRC)"),
    (1, "Written informed consent in Urdu | Right to withdraw at any time"),
    (1, "All lab results shared with participants & treating physicians"),
    (1, "Data anonymised using study IDs; stored on password-protected servers"),
])


# ═══════════════════════════════════════════════════════════════════════════════
# SLIDE 11 — Strengths & Limitations
# ═══════════════════════════════════════════════════════════════════════════════
slide = prs.slides.add_slide(blank)
add_rect(slide, 0, 0, 13.333, 7.5, LIGHT_GREY)
add_rect(slide, 0, 0, 13.333, 1.15, DEEP_NAVY)
add_rect(slide, 0, 1.15, 13.333, 0.07, DARK_RED)
add_text(slide, "Strengths & Limitations", 0.35, 0.18, 12.5, 0.85, size=28, bold=True, color=WHITE)

# Two column layout
# Strengths column
add_rect(slide, 0.3, 1.35, 6.0, 5.9, MID_BLUE)
add_text(slide, "✔  STRENGTHS", 0.45, 1.42, 5.7, 0.5, size=17, bold=True, color=WHITE)
strengths = [
    "Faster & cheaper than cohort studies",
    "No need to wait for disease to develop",
    "Examines multiple risk factors at once",
    "Multi-centre design — broad generalisability",
    "Includes Pakistan-specific exposures (naswar)",
    "Incident cases — avoids survivor bias",
    "Large sample size (900+ cases, 900+ controls)",
]
y = 2.0
for s in strengths:
    add_text(slide, "• " + s, 0.45, y, 5.7, 0.48, size=14, color=WHITE)
    y += 0.52

# Limitations column
add_rect(slide, 6.9, 1.35, 6.1, 5.9, WHITE)
add_text(slide, "✘  LIMITATIONS", 7.05, 1.42, 5.8, 0.5, size=17, bold=True, color=DARK_RED)
limitations = [
    "Cannot calculate incidence rates directly",
    "OR approximates RR only when disease is rare",
    "Retrospective exposure — recall bias risk",
    "Hospital controls may share some risk factors",
    "Temporal sequence not always definitive",
    "Selection of truly representative controls is challenging",
]
y = 2.0
for l in limitations:
    add_text(slide, "• " + l, 7.05, y, 5.8, 0.48, size=14, color=DARK_GREY)
    y += 0.52


# ═══════════════════════════════════════════════════════════════════════════════
# SLIDE 12 — Summary & Conclusions
# ═══════════════════════════════════════════════════════════════════════════════
slide = prs.slides.add_slide(blank)
add_bullet_slide(slide, "Summary & Expected Impact", [
    (0, "Study Summary"),
    (1, "Hospital-based case-control study | ~900 cases + ~900 controls | 3 cities"),
    (1, "18-month timeline | Ethics-approved | Multi-disciplinary team"),
    (0, "Key Expected Findings"),
    (1, "Hypertension, dyslipidaemia, smoking and diabetes are the dominant ORs"),
    (1, "Abdominal obesity and physical inactivity are independent contributors"),
    (1, "Smokeless tobacco (naswar) will emerge as a locally-specific risk factor"),
    (0, "Policy Impact"),
    (1, "PAR estimates will directly guide national CVD prevention priorities"),
    (1, "Supports blood pressure screening, anti-smoking campaigns, and dietary education"),
    (1, "Findings to be published in JPMA & Eastern Mediterranean Health Journal"),
    (1, "Shared with Pakistan Ministry of National Health Services"),
])


# ═══════════════════════════════════════════════════════════════════════════════
# SLIDE 13 — Thank You
# ═══════════════════════════════════════════════════════════════════════════════
slide = prs.slides.add_slide(blank)
add_rect(slide, 0, 0, 13.333, 7.5, DEEP_NAVY)
add_rect(slide, 0, 0, 13.333, 0.18, DARK_RED)
add_rect(slide, 0, 7.22, 13.333, 0.28, DARK_RED)
add_rect(slide, 0, 0.18, 0.12, 7.04, MID_BLUE)

add_text(slide, "Thank You", 0.5, 1.5, 12.3, 1.5, size=54, bold=True, color=WHITE, align=PP_ALIGN.CENTER)
add_rect(slide, 2.5, 3.2, 8.333, 0.05, LIGHT_BLUE)

add_text(slide, "Questions & Discussion", 0.5, 3.4, 12.3, 0.8,
         size=24, italic=True, color=LIGHT_BLUE, align=PP_ALIGN.CENTER)

add_rect(slide, 3.5, 4.4, 6.3, 1.85, MID_BLUE)
add_text(slide, "SHAHAB", 3.6, 4.5, 6.1, 0.65, size=28, bold=True, color=WHITE, align=PP_ALIGN.CENTER)
add_text(slide, "Group: Med21C", 3.6, 5.15, 6.1, 0.45, size=20, color=LIGHT_BLUE, align=PP_ALIGN.CENTER)
add_text(slide, "July 2026", 3.6, 5.65, 6.1, 0.45, size=16, italic=True, color=LIGHT_BLUE, align=PP_ALIGN.CENTER)


# ── Save ─────────────────────────────────────────────────────────────────────
out_path = "/tmp/workspace/cvd-pakistan-pptx/CVD_Pakistan_CaseControl_SHAHAB_Med21C.pptx"
prs.save(out_path)
print(f"Saved: {out_path}")

Running Command

cd /tmp/workspace/cvd-pakistan-pptx && python create_pptx.py

Running Command

python3 -c " from pptx import Presentation prs = Presentation('/tmp/workspace/cvd-pakistan-pptx/CVD_Pakistan_CaseControl_SHAHAB_Med21C.pptx') print(f'Total slides: {len(prs.slides)}') for i, slide in enumerate(prs.slides, 1): texts = [s.text_frame.text.strip()[:60] for s in slide.shapes if s.has_text_frame and s.text_frame.text.strip()] print(f' Slide {i}: {texts[0] if texts else \"(no text)\"}') "

Your presentation is ready for download:

CVD Pakistan CaseControl SHAHAB Med21C

Presentation · PPTX

Presentation Overview — 13 slides, 16:9 widescreen, professional navy/red theme:

#	Slide Title
1	Title Slide — Name: SHAHAB, Group: Med21C (prominently displayed)
2	Background & Rationale
3	Study Objectives
4	Study Design Overview (sites, sample size, timeline)
5	Case-Control Conceptual Framework + 2×2 table with OR formula
6	Selection of Cases & Controls (inclusion/exclusion criteria)
7	Data Required (sociodemographic, clinical, biochemical, lifestyle)
8	Hypothesised Risk Factors & Expected Odds Ratios (table format)
9	Methods of Analysis (χ², t-test, logistic regression, PAR%)
10	Bias Control & Ethical Considerations
11	Strengths & Limitations (two-column layout)
12	Summary & Expected Policy Impact
13	Thank You / Discussion

This is a shared conversation. Sign in to Orris to start your own chat.