Incidence of Postoperative Cognitive Dysfunction in Older Women Undergoing Pelvic Organ Prolapse Surgery

Abstract

Introduction and Hypothesis:

Postoperative cognitive dysfunction (POCD), a transient impairment of memory, concentration, and information processing, has been reported after 7–26% of non-cardiac surgeries with associated increase in morbidity and death. Our primary aim was to determine the incidence of POCD two weeks after prolapse surgery in women ≥60 years. Our secondary aim was to identify risk factors for POCD.

Methods:

Prospective cohort study of women ≥ 60 years old scheduled for pelvic organ prolapse surgery. Exclusion criteria included cognitive impairment history, major neurologic disorder, and abnormal cognition screen. A comprehensive neuropsychological (NP) battery (eight tests), administered two weeks pre- and post-surgery assessed premorbid IQ and domains of attention, memory, and executive function. The primary outcome was defined as decline of ≥1 SD on ≥ 2 NP tests or decline of ≥ 2 SD on ≥ 1 test. Raw scores were transformed to Z-scores.

Results:

NP testing was completed by 72 women, median age 72 (IQR 69–77) years. Procedures included 16 (22.9%) laparoscopic sacrocolpopexies, 23 (32.9%) transvaginal reconstructions and 29 (41.4%) obliterative surgeries, performed under general (63, 90%), regional (5, 7.1%), or sedation (2, 2.9%) anesthesia with a median hospital stay of 0.6 (IQR 0.6–0.75) days. POCD incidence was 33.3% (n=24). POCD was associated with greater frailty (p=0.006), and higher baseline depression (p=0.05) but not with older age (p=0.77) nor inhalational gas use (p=1.0).

Conclusion:

In this cohort, one in three women manifested POCD two weeks after prolapse surgery. Preoperative counseling should include discussions on POCD given its detrimental impact on postoperative recovery and independence.

Brief Summary:

One-third of women had postoperative cognitive dysfunction (POCD) two weeks after prolapse surgery. Increased POCD awareness is needed to inform surgical decision-making and recovery planning.

Introduction:

Postoperative cognitive dysfunction (POCD) is primarily a transient deterioration of cognition after surgery characterized by impairment of memory, concentration and information processing speed. It is distinct from delirium and dementia. POCD typically lasts days to months, however, has been reported to last up to 7.5 years [1]. Despite variability in its duration, POCD in older adults (≥60 years) is associated with delayed postoperative recovery, greater loss of independence, increased health care costs, and increased morbidity and mortality [1–4]. Health care professionals often underestimate the incidence and negative influence of POCD as the diagnosis requires neuropsychological testing, which is not routinely performed in the clinical setting.

Pelvic organ prolapse (POP) surgery is a common and increasing elective surgery performed in women over age 60. The demand for care of pelvic floor disorders, like pelvic organ prolapse, is projected to increase by 35% between 2010 and 2030 given the anticipated 24% growth of the US population [5]. Thus, the patient requests for POP surgery will continue to increase over the next few decades as the US population of older adults grows.

POCD after elective, non-cardiac surgery ranges from 7% to 26% [1–4]. One of the largest studies assessing POCD after non-cardiac surgery reported an incidence of 25.8% at one week postop and a POCD prevalence of 9.9% at 3 months postoperatively [4]. Risk factors for POCD are important to identify in the urogynecologic patient population given their advanced age, a known risk factor for POCD development in elective, noncardiac surgery patient populations [1, 3, 4], and the availability of nonsurgical management options for many women. It is imperative that we better understand the risk of POCD and its impact on postoperative recovery in the urogynecologic patient population as the potential for this outcome may influence the decision to proceed with surgery.

The incidence of POCD in the urogynecologic patient population has not been well established. Given this paucity of research, we aimed to determine the incidence of POCD two weeks after POP surgery in women 60 years and older. We hypothesized that the incidence rate is similar to reported rates of 20% in populations of older persons undergoing major elective, non-cardiac surgery. Our secondary aim was to identify factors associated with POCD at 2 weeks after surgery. We hypothesized that older age, use of inhalational anesthetic agents and frailty would be associated with POCD 2 weeks after surgery.

Materials and Methods:

Participants and Exclusion Criteria:

Study of cognitive impairment in older women after pelvic organ prolapse surgery (SCOOP) is a prospective cohort study of women ≥ 60 years who were recruited by 6 urogynecologic surgeons at an academic medical center. This study was performed with IRB study approval from the University of Pittsburgh (IRB#:STUDY19020276). Women were identified for enrollment after surgical orders for major pelvic organ prolapse repair were placed. Women were excluded if they reported any of the following: 1) diagnosis of dementia and/or cognitive impairment assigned by their physician/provider, 2) inability to speak or read English, 3) alcohol or drug abuse, 4) severe hearing impairment despite corrective interventions, 5) severe vision impairment despite corrective interventions, 6) current use of antipsychotic medication for psychiatric diagnosis, and 7) history of cerebrovascular accident or major neurologic disorder. To ensure the absence of undiagnosed cognitive impairment at baseline, consented participants were further screened with the Modified Mini Mental State Exam (3MS). Those who scored <84 were excluded from participation [6].

Measures and Instruments:

A baseline assessment, inclusive of participant interview, chart review, frailty assessment and neuropsychological testing, was conducted in person by appointment within 4 weeks of scheduled surgery. This interval was selected to coincide with routine preoperative appointments and to limit the period for occurrence of potential confounders such as occult cerebral infarctions.

Participant Interview/Chart Review:

Each participant was interviewed to ascertain clinical and sociodemographic information. Table 1 summarizes these variables. Medication information was verified through chart review of each participant’s electronic medical record.

Table 1.

Baseline and Outcome Variables with Ascertainment Method

Data Collection Method	Variables
Patient Interview	Self-reported hearing impairment, self-reported visual impairment, substance use, medication reconciliation, history of psychiatric disorders, education level, income level, employment/job history, number of dependents, home description, fall history, number of persons who live with subject, use of assistive ambulation device
Chart Review	Age, race/ethnicity, body mass index, gravidity, parity, medication history, medical co-morbidities, surgical history, psychiatric history, and pelvic organ prolapse quantification stage
Self-reported validated questionnaire scores	Beck Anxiety Index (anxiety), Geriatric Depression Scale-SF (depression), Numerical Pain Rating Scale (pain), Pelvic Floor Distress Inventory-20 (pelvic floor disorder symptoms)
Frailty Assessment	Frailty will be measures using the Fried Frailty Index  • Components of frailty index: weight loss, decreased hand grip strength, exhaustion, slowed walking speed, low activity  • Each criterion given a score of 0 or 1  • Total score classification: 0 Not frail, 1–2 intermediately frail; 3–5 frail

Validated Questionnaires and Scales:

Participants were instructed to complete three validated self-report questionnaires during their baseline assessment. These measures included the Beck Anxiety Inventory (BAI) [7], the Geriatric Depression Scale-Short Form (GDS-SF) [8], and the Numerical Pain Rating Scale (NPRS) [9]. We elected to measure these quality of life specific symptoms due to the known association between cognitive impairment and anxiety, depression and pain symptoms [7, 10–12]. Further, these questionnaires have been used extensively in populations of older adults and have good psychometric properties [9, 13–14]. Medication use history was used to calculate Anticholinergic Cognitive Burden (ACB) Scale score for each participant. We chose to evaluate the association between ACB scores and POCD given the known association of increased anticholinergic activity and cognitive impairment [15].

Frailty Assessment:

Fried frailty index [16] was calculated for each participant. The 5 components assessed for this index include self-reported unintentional weight loss, weakness (measured by hand grip strength), self-reported exhaustion, slow walking speed, and low physical activity. Intermediate frailty or pre-frail status was defined as a syndrome in which 1 or 2 of these criteria were met. Frailty was defined as a clinical syndrome in which 3 or more of these criteria were present [16].

Neuropsychological Testing:

Baseline cognitive function was evaluated through performance on a comprehensive battery of neurocognitive tests that assesses general mental status and includes multiple measures of major domains of cognitive functioning, including attention, memory, and executive functions. The battery is consistent with current conceptualizations of brain-behavior relationships. Each of the 8 tests has demonstrated reliability and validity [17]. Premorbid intellectual ability was assessed with the Wechsler Test of Adult Reading (WTAR). Measures of delayed episodic memory included free recall and recognition of the California Verbal Learning Test-2 (CVLT-2) and the Brief Visuospatial Memory Test-Revised (BVMT-R). Measures of attention/processing speed included the Digit Symbol Substitution Test (DSST) and Trail Making Test Part A (TRL-A). Measures of executive functioning included Trail Making Test Part B (TRL-B), and Stroop Color-Word Inhibition Test (Stroop). The battery was administered by trained research staff under the supervision of a senior neuropsychologist (MAB). Testing took approximately 60 minutes to complete, and it incorporated rigorous procedures including counterbalanced alternate forms and masked assessors. Testing was performed at the University of Pittsburgh Medical Center offices or at the participant’s home depending on participant preference.

Surgery and Postoperative Assessment:

Participants underwent surgery as planned by the attending surgeon. Clinical details regarding the surgery, anesthesia administered, and the hospitalization were obtained from the electronic medical record. Variables abstracted included: surgical procedure and anesthesia type and duration, type and amounts of administered anesthetic medications, intraoperative complications, length of hospitalization stay, and disposition at discharge.

During the first week after surgery, women were screened for delirium due to its distinct characterization yet overlapping clinical manifestations with POCD. We screened women on post-operative days 0, 1, 3, 5 and 7, either in person or by telephone using the Confusion Assessment Method (CAM) [18]. This validated tool is used to assess features of delirium including acute onset, fluctuating course, inattention, disorganized thinking and altered level of consciousness, and it has been validated for in-person and telephone administration [19, 20].

Neuropsychological testing and the battery of quality of life/ symptom severity questionnaires were repeated 2 weeks +/− 2 days after the surgery. To minimize the impact of practice effect, alternate forms of neuropsychological tests were used as available. Postoperative clinical course was evaluated through review of the electronic medical record and confirmatory discussion with each participant at postoperative follow up visits. Data collected includes postoperative complications (as defined by Clavien-Dindo criteria [21]), home health nursing service utilization, postoperative emergency room visitation, and hospital readmissions. Postoperative clinical outcomes that occurred up to 6 weeks after surgery were included in the assessment.

Statistical Analysis Plan:

The primary outcome of POCD was defined as a pre-to-post decline of ≥ 1 standard deviation (SD) on ≥ 2 neurocognitive tests or a decline of ≥ 2 SDs on ≥ 1 test. This definition was selected based on previously used POCD definitions in published research on large cohorts examining POCD using neurocognitive testing batteries. The incidence rate for POCD was compared to a hypothesized rate of 20% using a one-sample test of proportions. This hypothesized rate was an estimation based on reported rates of POCD after major elective noncardiac surgery [1–4]. Chi-square (Fisher’s exact) and Student’s T (Mann U Whitney) tests assessed the association between POCD group and categorical or continuous sociodemographic/clinical factors. Our sample size of 70 subjects afforded the ability to estimate a margin-of-error of approximately 0.094 for our 95% confidence interval.

Results:

Of the 452 women who met eligibility criteria during the study period, 72 of 91 consented participants provided the completed set of pre- and postoperative neuropsychological test results. The remaining 19 women did not complete testing (n=8) or were withdrawn from the study (n=11). The 2 participants withdrawn by the principal investigator were due to an abnormal cognition screen at baseline and discovery of antipsychotic medication use after enrollment (Figure 1).

Figure 1.

Flow diagram of study participants

The median (interquartile range (IQR)) age was 72 (69–77) years. The median (IQR) number of years of education was of 13 (12–16). The majority of women were white (n=66, 91.7%) and almost one quarter of women reported living alone (n=16, 22.2%). Over half of the women in the cohort (n=47, 65.3%) were prefrail/intermediately frail by Fried Frailty Index criteria. Table 2 further summarizes demographics and clinical characteristics of the cohort. Standardized mean (SD) premorbid IQ scores were average to above average (106.03 (12.64)). Standardized scores of memory and executive functioning domain tests were average to above average, and attention test scores were notably average (Table 2).

Table 2.

Baseline Demographics, Clinical Characteristics and Standardized Neuropsychological Test Results

Demographic and Clinical Characteristics	All Participants (n=72)
Age, years^	72 (69, 77)
Race, white¥	66 (91.7)
Body Mass Index, kg/m2+	27.88 (25.07, 31.00)
Education, years^	13 (12, 16)
Fried Frailty Status¥
None	16 (22.2)
Prefrail/Intermediate	47 (65.3)
Frail	9 (12.5)
Subjective Hearing Loss¥	25 (34.7)
Subjective Vision Impairment¥	41 (56.9)
Smoking Status¥
Current	3 (4.2)
Former	23 (31.9)
Never	46 (63.9)
Living Alone¥	16 (22.2)
History of Depression	10 (13.9)
History of Generalized Anxiety Disorder	13 (18.1)
Baseline Geriatric Depression Scale Score, (range 0–15)	1.9 (2.2)
Baseline Beck Anxiety Index Score, (range 0–22)	5.3 (5.0)
Cognitive Domain	Baseline Testing Score* (n=72)
Measures of Global Cognition and Premorbid Intellectual Ability
Modified Mini-Mental Status Exam (3MS)	94.64 (3.87)
Wechsler Test of Adult Reading (WTAR)	106.03 (12.64)
Memory
CVLT-II List A Trials 1–5 T-Score	53.93 (9.45)
CVLT-II Short Delay Free Recall Z score	0.17 (0.93)
CVLT-II Long Delay Free Recall Z score	0.01 (0.94)
CVLT-II Discriminability Z score	0.04 (1.06)
BVMT-R Total Recall T Score	44.58 (12.29)
BVMT-R Delayed Recall T Score	48.24 (11.95)
Executive Functioning
DKEFS Verbal Fluency Letter Fluency Total Correct Scaled Score Published Norm	10.23 (3.25)
DKEFS Verbal Fluency Category Fluency Total Correct Scaled Score Published Norm	12.44 (3.21)
DKEFS Verbal Fluency Category Switching Total Correct Scaled Score Published Norm	11.55 (3.85)
Attention
WAIS Digit Span Total Scaled Score	10.40 (2.70)

^{^}Data presented as median (interquartile range)

^¥Data presented as n (%)

CVLT-II: California Verbal Learning Test – Version II

BVMT-R: Brief Visuospatial Memory Test – Revised

DKEFS: Delis-Kaplan Executive Functioning System

WAIS: Wechsler Adult Intelligence Scale

^*Data presented as mean (standard deviation)

Score range for standardized executive functioning tests: 1–19

Score range for standardized attention tests: 0 –19

Pelvic organ prolapse surgeries performed included 19 (26.4%) laparoscopic sacrocolpopexies, 23 (31.9%) transvaginal reconstructions and 30 (41.7%) obliterative vaginal surgeries. Of these surgeries, the majority (64 (88.8%)) were performed under general anesthesia. The incidence of POCD two weeks after pelvic organ prolapse surgery was 33.3% (n=24), significantly greater than the 20% reported in the literature (p=0.02). Twenty-three of the 24 women had at least 1 test with 2 SDs decline. The memory domain was most often affected (13/24 women) compared to executive functioning (11/24 women) and attention (11/24 women) domains. Additionally, 11 of the 24 women met POCD criteria with a ≥ 2 SD decline on 1 test. (see Figure 2). Only those women who took opioids chronically prior to surgery (n=3) reported taking opioids at 2-week postoperative neurocognitive testing.

Figure 2.

Cognitive Testing Changes by Domain and Participant

Tables 3 and 4 summarize the demographic, clinical characteristics, and perioperative factors by POCD group. Neither age nor use of inhalational anesthetic agents was associated with POCD development (p=0.77, p=1.0, CI [−0.2, 0.2] respectively). Development of postoperative delirium was not associated with subsequent manifestation of POCD (p=0.72, CI [−0.22, 0.13]). Amongst those with a positive screen for delirium, we observed resolution (negative CAM screen) within 7-days after surgery. All participants had a negative delirium screen on their last day undergoing CAM screening during postoperative week one. POCD was not associated with adverse events (p=0.76, CI [−0.28, 0.11]) nor serious adverse events (p=1.0, CI [−0.02, 0.06]) as categorized by Clavien-Dindo criteria. POCD was not associated frail status (p=0.07), but it was associated with higher Fried Frailty Index scores (p=0.006, CI [−1.19, −0.21]). There is an association between POCD and higher baseline geriatric depression scale scores (p=0.05, CI [−2.32, −0.28]).

Table 3.

Participant Demographics and Clinical Characteristics by POCD status

	All Participants (n = 72)	Participants without POCD (n=48)	Participants with POCD (n = 24)	p-value
Age, years				0.77
60 – 69 years	19 (26.4)	14 (29.2)	5 (20.8)
70 – 79 years	43 (59.7)	28 (58.3)	15 (62.5)
≥ 80 years	10 (13.9)	6 (12.5)	4 (16.7)
Race, white	66 (91.7)	43 (89.6)	23 (95.8)	0.66
Education, years				0.21
≤ 12 years	25 (34.7)	14 (29.2)	11 (45.8)
> 12 years to ≤ 16 years	38 (52.8)	26 (54.2)	12 (50.0)
> 16 years	9 (12.5)	8 (16.7)	1 (4.2)
Modified Mini-Mental Status Exam Score (range: 0–100)	94.6 (3.9)	94.5 (3.9)	94.8 (3.8)	0.77
ASA Classification				0.82
1	2 (2.8)	1 (2.1)	1 (4.2)
2	40 (55.6)	26 (54.2)	14 (58.3)
3	30 (41.7)	21 (43.8)	9 (37.5)
Frailty Status				0.07
Not Frail	16 (22.2)	14 (29.2)	2 (8.3)
Prefrail/Intermediate	47 (65.3)	30 (62.5)	17 (70.8)
Frail	9 (12.5)	4 (8.3)	5 (20.8)
Fried Frailty Index Score, (range 0–5)	1.2 (1.1)	1.0 (0.9)	1.7 (1.2)	0.006
Subjective Hearing Loss	25 (34.7)	16 (33.3)	9 (37.5)	0.73
Subjective Vision Loss	41 (56.9)	29 (60.4)	12 (50)	0.40
History of Depression	10 (13.9)	7 (14.6)	3 (12.5)	1.0
Baseline Geriatric Depression Scale Score, (range 0–15)	1.9 (2.2)	1.4 (1.6)	2.7 (2.8)	0.05
History of Generalized Anxiety Disorder	13 (18.1)	7 (14.6)	6 (25.0)	0.28
Baseline Beck Anxiety Index Score, (range 0–22)	5.3 (5.0)	5.1 (4.5)	5.8 (6.0)	0.94
Anticholinergic Cognitive Burden Scale Score	0 (0–1)	0 (0–1)	0 (0–1.5)	0.35

ASA: American Society of Anesthesiologists

Data presented as n(%), mean (SD), median (IQR)

Table 4.

Intraoperative and Postoperative Factors by POCD status

	All Participants (n=72)	Participants without POCD (n = 48)	Participants with POCD (n=24)	p-value
Surgical Approach^				0.45
Vaginal	53 (73.6)	34 (70.8)	19 (79.2)
Laparoscopic	19 (26.4)	14 (29.2)	5 (20.8)
Concomitant Hysterectomy^				1.0
Performed	42 (58.3)	28 (58.3)	14 (58.3)
Not Performed	30 (41.7)	20 (41.7)	10 (41.7)
Anesthesia Time^				0.62
Less than 3 hours	33 (45.8)	23 (47.9)	10 (41.7)
Greater than 3 hours	39 (54.2)	25 (52.1)	14 (58.3)
Length of Stay*, days	0.70 (0.60, 0.75)	0.65 (0.60, 0.70)	0.70 (0.60, 0.80)	0.33
Estimated Blood Loss*, mL	50 (30, 100)	50 (25, 100)	75 (45, 100)	0.27
Anesthesia Route^				0.57
Local	2 (2.8)	2 (4.2)	0 (0)
Regional	6 (8.3)	5 (10.4)	1 (4.2)
General	64 (88.9)	41 (85.4)	23 (95.8)
Use of Inhalational Gas^	15 (20.8)	10 (20.8)	5 (20.8)	1.0
Positive Postoperative Delirium Screening^	10 (13.8)	6 (12.5)	4 (16.7)	0.72
Adverse Events^#	56 (77.8)	36 (76.6)	20 (83.3)	0.76
Serious Adverse Events^#	1 (1.4)	1 (2.8)	0 (0)	1.0

^{^}Data presented as n(%)

^*Data presented as median (interquartile range)

^#Defined by Clavien-Dindo criteria; Serious Adverse Events include Clavien-Dindo category III, IV, and V

Discussion:

In this cohort of older women who underwent surgery for symptomatic pelvic organ prolapse, the incidence of POCD two weeks after surgery was 33.3%, which was significantly greater than the anticipated incidence of 20%. This greater incidence is likely related to the older age, median age 72 years, of our study cohort compared to prior studies of men and women with age means or medians of 60+ years [3]. Our study findings emulate those of Moller et al. who reported a 29% incidence of POCD in their large prospective study of 425 adults aged ≥70 years one week after abdominal, thoracic, or orthopedic surgery [4]. Monk et al. reported an POCD incidence of 41.4% one week after abdominal, thoracic, or orthopedic surgery in a cohort of 333 older adults aged ≥60 years [2]. Another factor contributing to the difference in our reported rate of POCD compared to incidence rates from other studies is the heterogeneity of the POCD criteria used in published research studies [1, 22]. Review of prior POCD research studies demonstrates wide variation in type of neuropsychological tests used to assess cognition, incorporation of control groups, and the statistical analysis criteria utilized to diagnose POCD [1, 3, 22]. For this study, we used a comprehensive neuropsychological test battery (similar to other large POCD studies), a POCD definition with criteria similar to other large studies, and test administration by trained neuropsychological assessors to ensure accurate assessment and to minimize ceiling effect. While it is possible our finding of a 33.3% POCD incidence is an overestimate, it likely to be accurate as the test battery used has the ability to appreciate very subtle declines in cognition that may ultimately impact functional status, postoperative recovery and well-being for older women. It is possible that we may have found a higher incidence of POCD if we tested 1 week rather than 2 weeks postoperatively. However, we were most interested in the 2-week postoperative timepoint given its clinical relevance to recovery and early independence for surgical patients after prolapse surgery. To our knowledge this is one of the first studies to examine POCD in this specific population of older women who underwent elective prolapse repair surgery.

We also found greater Fried Frailty index scores to be associated with POCD development in this cohort, but not frail status, most likely due to the small number of women in this cohort who met criteria (n=9). Preoperative frailty has been found to be associated with increased adverse outcomes and postoperative complications generally [16,23]. There are few studies that have reported a specific association between preoperative frailty and POCD development [1, 24]. Rothrock et. al performed a prospective cohort study of 100 older adults ≥65 years who underwent elective spinal surgery and found that 50% of those who were frail had persistent cognitive decline compared to their baseline. Notably, this study utilized a quality of life specific questionnaire (Post-operative Quality of Recovery Scale) to measure cognitive recovery [25]. This association is plausible given the vulnerability of older, frail persons who may subsequently have declines in organ functions that can impact their tolerance of the surgical experience and anesthesia. With this, frailty is a risk factor for POCD that should be considered in surgical decision making.

Baseline Geriatric Depression Scale scores were significantly higher in women with POCD. However, the clinical relevance is uncertain given the low absolute raw scores (score mean(SD) of 1.4(1.6) in women without POCD vs. 2.7(2.8) in women with POCD). The mean GDS-SF scores in both groups are below that of > 5 which has been associated with moderate or severe depressive symptoms. Current published literature includes differing reports of the relationship between depression and POCD. Monk et al. reported that preoperative history of depression was not associated with POCD in a cohort who underwent elective, noncardiac surgery [2]. Other literature has reported an association between depression and cognitive decline, outside of the surgical setting [11, 26] as well as associations between depression and POCD [27, 28]. History of clinical depression was not associated with POCD in this cohort, but the association between higher GDS-SF scores and POCD was statistically significant. This is an important finding and consideration given the greater prevalence of subthreshold depression compared to major depression amongst older women. Both are associated with decreased quality of life and increased morbidity [29].

We hypothesized that older age would be associated with POCD given that most studies on POCD have reported older age as an associated or predictive factor [1–4]. We also hypothesized that inhalational anesthetic use would be associated with POCD. A Cochrane Review by Miller et al. based on 28 studies reported that there was low-certainty evidence that propofol-based total intravenous anesthesia may reduce postoperative cognitive dysfunction (POCD) compared to inhalational anesthesia [30]. These factors were not associated with POCD development in our study cohort though we concede we are underpowered to inform on specific risk factors for POCD as this was an exploratory aim.

There are limitations of this study that should be considered with interpretation of our findings. First, we did not have a control group. We used baseline performance as a comparator to measure cognitive change. Ideally, inclusion of a control group, specifically with two assessments at the same interval of time (approximately four weeks) without the surgical intervention, would be ideal to allow for calculation of a reliable change index (RCI), which would account for ceiling/floor effects as well as test-retest reliability. We did not incorporate a control group into our study design given the feasibility challenges of identifying women who were offered surgical prolapse repair and had time to schedule two testing sessions with a four week interval prior to undergoing the surgical intervention. Second, though the age range and the spectrum of prolapse procedures performed generally reflect the population of women who seek surgical intervention for prolapse [31], this study’s predominantly white and regionally based (all from western Pennsylvania) cohort limits generalizability to other populations. Exclusion of women with preoperative cognitive impairment, a known risk factor for POCD development, also limits generalizability. However, we intentionally included preoperative cognitive impairment as an exclusion criterion as we sought to identify new cases of cognitive impairment related to the surgery. Third, a 2-week postoperative assessment timepoint may be considered too short. We purposely selected a 2-week postoperative assessment as this time best reflects when the acute anesthesia effects are resolved, patients are back at home (having resumed most functional activities of daily living), and minimal, if any, opioid medications are being used. It is also the timepoint when most safety-net resources such as home health nursing or aids are discontinued. Evered, et al. with the Nomenclature Consensus Workgroup has published recommendations for standardized nomenclature and definitions used in research on perioperative neurocognitive disorders, including postoperative cognitive dysfunction [22]. Given the significant heterogeneity in the definitions and measures used to assess postoperative cognitive dysfunction and other perioperative neurocognitive disorders, these recommendations will help with interpretation of results in this area of science. Based on these recommendations, the nomenclature for what was previously called POCD at a 2-week interval after surgery should now be referred to as ‘delayed neurocognitive recovery’ [22]. Although this study did not follow this particular cohort to time intervals greater than 2-weeks after surgery, the findings are still significant as these cognitive changes may impact the ability of women to care for themselves during the vulnerable period of postoperative recovery.

Strengths of this study included its prospective design and use of a comprehensive neuropsychological test battery by administered by trained assessors. Common cognitive tests, such as the 3MS or mini-cog, alone are not as sensitive in assessing cognitive decline as it relates to multiple cognitive domains. Awareness of POCD as a potential complication is essential given its impact on surgical decision making and informed consent discussions between surgeons and patients, especially in the setting of elective surgery performed primarily for quality of life indications.

POCD is a common occurrence amongst older women undergoing surgical repair of pelvic organ prolapse. Our study confirmed that as many as 33% of women demonstrated persistent deterioration from their baseline in memory, executive function, and/or attention domains two weeks after surgery. This analysis identified greater depression and frailty scores as risk factors for POCD development. Surgeons, patients and their families should be aware of this diagnosis and its associated risk factors to inform surgical decision making and postoperative recovery planning. Our findings have implications for other surgical specialties that perform major elective noncardiac procedures in women and warrants further investigation. Future studies should investigate the persistence of POCD over time and its impact on functional status and postoperative recovery.