Patient understanding and satisfaction in informed consent for total knee arthroplasty: A randomized study


  • Michael R. Johnson,

    1. University of Minnesota Medical School, Minneapolis
    Search for more papers by this author
  • Jasvinder A. Singh,

    1. Birmingham VA Medical Center and University of Alabama at Birmingham
    Search for more papers by this author
    • Dr. Singh has received consultant fees, speaking fees, and/or honoraria (less than $10,000 each) from URL Pharmaceuticals, Savient, Takeda, Novartis, and Abbott; research and travel grants from Takeda, Savient, Wyeth, Amgen, the Arthritis Foundation, University of Minnesota, Cochrane Library, and Department of Veterans Affairs; and investigator-initiator grants from Allergan, Savient, and Takeda.

  • Thomas Stewart,

    1. University of Minnesota Medical School, Minneapolis
    Search for more papers by this author
  • Terence J. Gioe

    Corresponding author
    1. Minneapolis VA Medical Center and University of Minnesota Medical School, Minneapolis
    • Minneapolis Veterans Affairs Medical Center, Section 112E, 1 Veterans Drive, Minneapolis, MN 55417
    Search for more papers by this author
    • Dr. Gioe has received research support (more than $10,000) from Depuy, Inc.

  • The views expressed in this article are those of the authors and do not necessarily reflect the position or policy of the Department of Veterans Affairs or the US Government.



Informed consent is a critical component of all surgical procedures, but patients' understanding and recall of the potential risks/benefits is poor. We hypothesized that utilization of multiple standardized education modalities in the informed consent process would allow for better retention and a more informed patient.


A total of 151 patients undergoing primary total knee arthroplasty (TKA) were randomized to 3 groups: group 1 received standardized informed consent and a paper handout detailing the risks/benefits of TKA; group 2 received standardized informed consent, a paper handout, and a video discussing the risks/benefits of TKA; and group 3 followed the same process as group 2 plus formal nurse education. All patients completed a 15-item questionnaire (risks, indications, and expectations) immediately following this consent process on the morning of surgery and 6 weeks postoperatively. We used t-test and analysis of variance for data analyses.


There was no difference (P = 0.79) in satisfaction with the consent process between the 3 groups; 92–97% of the patients rated the consent process as good to excellent at all time points. The number of correct answers did not differ significantly between the groups at any time period (P = 0.31–0.81). Scores dropped significantly (P = 0.004) from preoperatively to the 6-week postoperative visit in all groups combined. A higher level of satisfaction with the process was reflected in higher scores preoperatively in all groups (P = 0.028).


Preoperatively, patients satisfied with the consent process may have better recall of risks/benefits and expectations of surgery. Neither retention nor satisfaction was influenced by reinforcement methods, such as video or nurse education; they may therefore be unnecessary.


Many surgeons still view informed consent as a medicolegal construct designed to ensure that the patient understands the risks and benefits of surgery. Over time, informed consent has evolved from an imposition of the surgeon's will to a partnership between the surgeon and patient that allows for an informed and willing patient (1). Informed consent may now more properly be seen as a “surgical virtue” (2), a way to empower patients to make decisions about their health care even when they are physically unable to voice their opinion (e.g., when under general anesthesia).

Legally, informed consent must contain a discussion of the diagnosis and the proposed procedure, the alternative therapy, the risks and benefits of surgical intervention, and the prognosis if no intervention is pursued (3). In addition, the patient must have “decision-making capacity,” defined as a combination of understanding, appreciation, reasoning, and expression of choice (4). Since the latter 3 concepts are difficult to measure, most studies of informed consent have focused on recall of information presented as a substitute for understanding.

Retention of information is selective and surgical benefits are recalled much better than risks. The retention of information discussed during the surgical informed consent process is very low. One study of 100 patients showed that retention of information less than 1 week after consent was obtained was only 33% of the material presented (5). In another study of 100 patients questioned 2–5 days after surgery, 27% of patients did not know what organ had been operated on (6).

How can we as surgeons improve the patient's understanding and comprehension of the proposed procedure and its inherent risks and benefits? Prior efforts have generally focused on intensive education, written handouts, or video presentations (7–9). Another way to improve the informed consent process may be to standardize what patients are told and how the information is presented. The Minneapolis Veterans Affairs Medical Center utilizes a computerized consenting process with an electronic signature known as iMedConsent (Dialog Medical). This allows for a standardized template to discuss expected risks and benefits of surgery, and in theory lessens the possibility of omitting important information. Computerized consent has been validated in previous studies showing a more comprehensive consent process than standard paper forms (10).

A structured consent process, in combination with one or a number of the aforementioned modalities (nursing education, handout, video, etc.), may reasonably be expected to improve surgical informed consent. We hypothesized that overall patient knowledge/retention would be influenced by the type of consent process and that more intensive training and review would result in better patient retention and recall. We further hypothesized that the knowledge retention would decline following the consent process and surgery, but in a stratified manner based on the nature of preoperative teaching.

Significance & Innovations

  • A preoperative standardized informed consent along with a paper handout before total knee arthroplasty provides good communication of the risks/benefits of knee arthroplasty.

  • Preoperatively, patients satisfied with the consent process may have better recall of the risks/benefits and expectations of surgery.

  • Neither retention nor satisfaction was influenced by reinforcement methods, such as video or nurse education; they may therefore be unnecessary.


After appropriate approval through the Minneapolis Veterans Affairs Medical Center Institutional Review Board, a prestudy power analysis was done to determine appropriate study numbers. Assuming a beta error of 20% and an alpha error (P value) of 0.05, a difference of 2 “correct” answers to questions between groups, and an SE of 20%, the number of patients required in each of the proposed groups was 50. Assuming a 15% missing response/nonresponse rate, a total of 176 patients were consented for the study to account for those missing questionnaires. Inclusion criteria included patients undergoing primary total knee arthroplasty (TKA), English as a native language, and patients that made their own medical decisions. Patient education level (grade; high school graduation = 12) was also obtained.

A total of 176 patients were randomized to one of three groups based on the week in which they presented for their preoperative visit to finalize preparations for the TKA. This method of randomization was chosen to prevent confusion during a busy clinic schedule, to allow for appropriate scheduling of the video and nursing education, and to avoid contamination of the process by patient intercommunication. All patients in all groups received iMedConsent in the standard fashion from an orthopedic surgery resident (level G3 or above) supervised by an attending staff surgeon.

Group 1 received the iMedConsent and a customized written handout (see Supplementary Appendix A, available in the online version of this article at further detailing the risks, benefits, and expectations of TKA surgery. Group 2 received the iMedConsent, the aforementioned handout, and a standard video from the American Academy of Orthopedic Surgeons (“Total Knee Replacement: Improving Quality of Life”) further detailing the risks, benefits, and expectations of TKA surgery. Group 3 received the iMedConsent, the handout, the video, and a formal education session with our nurse coordinator/educator additionally reviewing the risks, benefits, and expectations of TKA.

Once patients completed their assigned consent process, they took a questionnaire designed to assess their understanding and recall of information just presented to them regarding the planned TKA (Figure 1). The questionnaire consisted of 14 multiple-choice questions and one open-ended question (see Supplementary Appendix B, available in the online version of this article at This questionnaire specifically underscores what we deemed to be of pivotal importance to the understanding of TKA surgery. The questionnaire targeted risks (3 questions), indications (2 questions), expectations (5 questions), and a broad understanding of the surgical procedure (3 questions). The questionnaire was submitted to the Veterans Affairs Medical Center Medical Education Department to assure the questions and answers were at a seventh-grade reading level. The questionnaire was readministered the day of surgery and again at 6 weeks post-TKA in the followup clinic. Questionnaires were collected immediately following administration.

Figure 1.

Effect of consent method on knowledge score at various time points. * = overall number of correct answers from a total of 13 knowledge questions.

Analysis of the summary descriptive statistics, including mean, median, range, proportion, and SDs for overall score at various time points for each group, was performed. Depending on the distribution of scores (normal versus not normal), we planned on using analysis of variance or a nonparametric test to compare the 3 groups. Since the variables were normally distributed, analysis of variance was used for comparisons of continuous variables. For comparison of knowledge scores by the patient's overall rating of the process as excellent or not, we used the independent-sample t-test (t-test is a special case of analysis of variance, when a variable has 2 categories). For categorical variables, we used the chi-square test. A P value less than 0.05 was considered significant. To evaluate patterns of knowledge retention and understanding, individual questions of interest were analyzed separately. Overall satisfaction with the consent process was correlated with individual question responses and the overall score using Spearman's correlation coefficients.


The patients were enrolled between April 1, 2008, and March 30, 2009, in a consecutive fashion. A total of 228 patients were eligible for the study and 52 declined to participate; of the 176 patients that were enrolled, 25 were missing at least one questionnaire. Standard demographics for each group are listed in Table 1. There was no significant difference between groups based on age, sex, and education level.

Table 1. Demographic characteristics of the study population
 Group 1: handout + consent (n = 49)Group 2: handout + consent + video (n = 50)Group 3: handout + consent + video + teaching (n = 52)PAll patients combined (n = 151)
Age, mean ± SD years68 ± 965 ± 967 ± 90.1667 ± 9
Education, mean ± SD years14 ± 313 ± 213 ± 20.04713 ± 2

All of the preoperative, operative day, and 6-week followup questionnaires were completed and returned for all 151 patients included in the study. The mean ± SD time between the preoperative and operative day assessments was 9.9 ± 5.6 days. The response rate was excellent, with very few missing answers (0–5%; for most questions 0–2%). We did see a greater number of missing answers early in the study; however, this was remedied by having our nurses check over the questionnaires upon completion to make sure no questions were skipped. We were unable to contact 3 patients to obtain education level data.

The questionnaire was broken down into categories for further analysis. Questions 1–3 involved indications and expected results of surgery. Questions 4–6 involved procedural questions with regard to the surgery and implants. Questions 7–9 looked at risks of surgery and questions 10–13 evaluated peri- and postoperative results. Although certain questions were answered correctly by a very high percentage, no significant categorical trends were noted among the group as a whole (Table 2).

Table 2. Questionnaire responses for each question at each time point*
 Preoperative, no./total (%) correctDay of surgery, no./total (%) correct6-week postoperative, no./total (%) correct
  • *

    Differences in denominators reflect some patients skipping/missing questions. Q14 (satisfaction): How would you rate the informed consent and educational process that you went through before your knee surgery? Q15 (open-ended question): What would you have liked to hear more about during the informed consent and educational process? TKA = total knee arthroplasty.

Indications and expectations   
 Q1. Knee replacement surgery is elective, that is, a surgery you choose to have. What other way could your knee pain be helped apart from surgery?115/146 (79)115/144 (80)115/142 (81)
 Q2. What other surgery can be done for knee pain caused by arthritis?99/136 (73)85/134 (63)77/130 (59)
 Q3. Which of the following is the MOST EXPECTED BENEFIT of total knee replacement surgery?110/149 (74)110/150 (73)110/147 (75)
Knowledge of implant and procedure   
 Q4. What is actually replaced during knee replacement surgery?112/141 (79)107/141 (76)105/137 (77)
 Q5. The knee replacement itself is made of:139/149 (93)132/148 (89)119/145 (82)
 Q6. It is important that you feel no pain during surgery. Which of the following are methods of anesthesia that might be used during your knee replacement surgery?132/148 (89)118/146 (81)110/142 (77)
Risks related to TKA surgery   
 Q7. No surgery is without risks. Which of the following complications can occur after knee replacement surgery?113/151 (75)121/149 (81)111/145 (77)
 Q8. Infection can occur after any operation. How likely is an infection in your knee following the operation?105/144 (73)94/145 (65)90/145 (62)
 Q9. Blood loss occurs with knee replacement surgery and you may require a transfusion. How likely is a serious illness as a direct result of receiving blood?117/144 (81)117/142 (82)108/141 (77)
Peri- and postoperative results   
 Q10. How long are patients TYPICALLY in the hospital for this operation?141/151 (93)146/150 (97)138/143 (97)
 Q11. How long will it be before the new knee replacement feels as good as it will ever get?119/149 (80)113/147 (77)127/149 (85)
 Q12. No surgery can guarantee results. Which of the following best describes the EXPECTED result from successful knee replacement surgery?137/150 (91)132/149 (89)121/148 (82)
 Q13. Because the knee replacement is a mechanical device, it can wear out or fail. How likely is your new knee replacement to last at least 10 years?118/148 (80)118/144 (82)119/149 (80)

For all 13 knowledge questions, the last multiple choice option was “Don't remember receiving this information,” which implied failure of communication. For various questions, this ranged from 0–15 patients (0–10%) at the preoperative assessment, from 1–17 patients (1–11%) at the operative assessment, and from 2–21 patients (1–14%) at the postoperative period. Therefore, with the exception of the 6-week followup, communication failure was reported by 11% or fewer patients.

Although in general scores decreased slightly from preoperative to 6-week followup in all groups (Table 3), and groups 2 and 3 showed a trend toward a slight improvement over group 1 in the preoperative (P = 0.11) and day of surgery (P = 0.08) testing, this was not significant. No difference (P = 0.80) was noted between the groups at the 6-week testing point.

Table 3. Effect of consent method on knowledge score at various time points
 Preoperative, mean ± SD*Day of surgery, mean ± SD*6-week postoperative, mean ± SD*
  • *

    Overall number of correct answers from a total of 13 knowledge questions.

Group 1: handout + consent10.1 ± 2.410.2 ± 2.310.5 ± 1.7
Group 2: handout + consent + video10.8 ± 1.610.3 ± 1.810.2 ± 2.7
Group 3: handout + consent + video + teaching11.1 ± 1.511.0 ± 1.210.4 ± 2.2
P by analysis of variance0.110.080.80

Overall, the consent process was rated favorably by the patients. The satisfaction rates did not differ significantly by consent group immediately after the consent (P = 0.85), operative day (P = 0.33), or 6 weeks after the surgery (P = 0.07) (Table 4). On the day of surgery, patients who rated the consent process as very good or excellent scored significantly (P = 0.02) higher in knowledge than those with lower ratings. This difference was not noted at the other time points, however (Table 5).

Table 4. Effect of consent method on overall rating of the consent process
 Excellent/very good rating of the informed consent and educational process, %*
PreoperativeDay of surgery6-week postoperative
  • *

    Reference categories: poor, fair, good, very good, and excellent overall rating of the process.

Group 1: handout + consent948677
Group 2: handout + consent + video949490
Group 3: handout + consent + video + teaching969292
Table 5. Association of patient's satisfaction rating with knowledge scores at various time points
Time of testingPoor, fair, or good rating, mean ± SDVery good or excellent rating, mean ± SDP
Preoperative9.0 ± 2.310.7 ± 1.90.07
Day of surgery9.3 ± 2.710.6 ± 1.60.02
6-week postoperative11.2 ± 2.110.4 ± 1.80.21

Level of education was evaluated as a variable that might affect retention and understanding. Although patients with less than an eighth-grade level of education had the lowest scores at all time periods, this was not significant (Table 6). Since there were very few patients with less than 12 years of education, a primary analysis compared patients with ≤12 years of education versus >12 years by type of consent. For patients in the former group, the type of consent process was not associated with the knowledge score at the preoperative (P = 0.78), operative (P = 0.36), or postoperative points (P = 0.25). The level of education did not seem to prevent decline of retention.

Table 6. Association of education with knowledge score at various time points
 Preoperative, mean ± SD*Day of surgery, mean ± SD*6-week postoperative, mean ± SD*
  • *

    Overall number of correct answers from a total of 13 knowledge questions.

Years of education   
 ≤88.7 ± 1.29.6 ± 1.19.0 ± 2.8
 9–1210.6 ± 2.010.3 ± 2.010.2 ± 2.8
 >1210.8 ± 1.910.8 ± 1.510.5 ± 1.6


The surgeon's goal during the frequent and often difficult task of informed consent is to present information about a proposed procedure in an impartial way to a patient who has the most to gain or lose from the outcome. Informed consent implies the patient has the decision-making capacity, education, appreciation of risk, and freedom of choice to consent to a procedure (4). At issue is the assumption that the patient has processed and understood the information posed to them. We examined 3 possible adjuncts to a standardized computerized consent process to determine which might be most helpful in maximizing patient knowledge and recall of a common orthopaedic procedure.

The use of a standardized computer-based consent process has shown to improve consistency in the consent process (10). Use of the iMedConsent form within the Veterans Administration, with language geared toward a seventh-grade education, allows a discussion in nonprofessional terms improving compliance and patient satisfaction (11). The high satisfaction rate with the informed consent process in all groups in our study might be directly attributed to the common thread of the standardized iMedConsent itself.

Historically, recall of information given during the informed consent process is extremely low. Robinson and Merav found that only 29% of information was retained following information given with regard to proposed thoracic surgery (12). Hutson and Blaha showed that only 25% of complications could be recalled in their total joint population (9). Finding an appropriate aid to the surgeon's discussion alone to maximize a patient's understanding of the proposed procedure and its risks and benefits has proven to be more difficult than one might envision.

The use of handouts in addition to the traditional consent process is widely utilized and has received some scrutiny. Patients randomized to receive written information in addition to oral communication were more likely to understand the risks of cosmetic (13), head and neck (14), and maxillofacial (15) surgery than those who received oral communication only. Turner and Williams looked specifically at total joint arthroplasty patients and found that even with a handout given after the consenting process, recall of complications related to TKA and total hip arthroplasty ranged from 0–61% at 6 weeks postoperatively (16). Langdon et al (7) noted a significant improvement in recall in patients receiving written as opposed to verbal information regarding hip arthroplasty, but recall of complications remained poor in both groups. Since handouts of some type are a common adjunct to the educational process for many surgeons, we chose to use a customized written handout in all groups as the “baseline.” Although group 1 scores in our study were the lowest of the 3 groups, the scores remained stable and were not significantly worse than the groups with adjunctive methods. Patients may refer back to the handout at future time points, and it may prove helpful in a medicolegal sense if documentation that the patient has received it is recorded (17).

Other adjuncts to the consent process that have been commonly employed include video and nurse educators. Rossi et al examined the use of video for arthroscopic procedures consent and found improved patient recall of information, especially in the less educated (8). In a recent study, the use of a multimedia information program improved patients' self-assessed sense of understanding of cholecystectomy, again most significantly aiding patients with a lower level of education (18). Kikuchi et al examined physician education followed by nurse education two months prior to cataract surgery. At the time of surgery, educated patients could recall only 50% of the information they had been taught (19). A study testing recall of patients undergoing total joint replacement showed that despite intense tutoring from a nurse educator, recall of risks at 6 months postoperatively ranged from 3–28% (9). Direct comparisons with our study are difficult given the different timeframes employed in these studies, but we could not demonstrate that the commonly employed adjuncts of video and/or nurse education improved recall significantly at any time point.

Education level plays a major role in the ability of patients to understand and recall the information presented to them, no matter how the information is presented. Lower education level has been shown to have a direct negative effect on patients' comprehension and memory (20). Rossi et al found that adjuncts such as a video in the informed consent process had a greater effect on improving understanding in patients with less than or equal to 12 years of formal education (8). Our study, however, does not show that level of education predicts a better understanding and recall of the consent information presented. In addition, patients with lower levels of education (≤12 years) did not perform better when the adjuncts of video or nurse teaching were added. However, due to the small number of patients with less than 12 years of education, the analysis was underpowered to find differences.

Our patients achieved 75–80% correct answers on key aspects regarding arthroplasty, which is excellent by most standards, but one can aim for an even higher target. A small proportion (<10%) of patients acknowledged concerns with doctor/patient communication during the consent process. It is possible that strategies to target patients who are likely to be at higher risk of poor communication should receive extra attention and a more detailed discussion. Our study could not find a difference based on education level, but other patient characteristics such as presence of depression or a poor social support network might represent the group(s) at higher risk.

The strengths of this study are its relatively uniform population undergoing a single common orthopedic procedure utilizing a standardized computerized consent document as the “baseline” for comparison. The study was well powered to assess patients' recall with regard to a 3-tiered adjunctive strategy for improving recall as part of the informed consent process. Among the study limitations is a 10% female population for a procedure, i.e., TKA, that generally has a female predominance, and potential bias due to the method of randomization. Although the Veterans Affairs population allows for a homogenous study population, we could not account for possible inherent male versus female differences in learning and retention. Certainly, the basic premise of utilizing recall as a surrogate for understanding must be seen as a limitation in a study of this type. However, many studies across multiple medical specialties have made this same assumption since there is no direct method of assessing “understanding.” We also utilized a nonvalidated questionnaire, although to our knowledge, there is no validated questionnaire that could be utilized for this particular study. We did produce a questionnaire that was understandable (i.e., seventh-grade reading level) and balanced with regard to testing risks, benefits, and postoperative outcomes and expectations. Finally, although the resident was supervised and followed the outline of the iMedConsent form, and staff was available for additional questions, it could be argued that a surgeon with greater experience might present the information to the patient in a more understandable or consistent fashion.

These findings have potential implications for referring physicians (primary care, family practice, rheumatologists, and others) as well. They should be reassured that the patients they refer for joint replacement surgery are undergoing an appropriate consent process with a standardized consent process and tailored handouts. Patients seek information regarding joint replacement surgery from peers, other physicians, surgeons, and internet resources, and may be relatively well informed about their joint replacement surgery. Referring physicians could also be provided with procedure-specific handouts tailored by their surgeon consultants, serving as a springboard for a more informed discussion when seeing the surgeon.

In conclusion, neither recall/retention nor satisfaction with the TKA consent process was significantly influenced by video or nurse instruction reinforcement methods; their routine use as adjuncts may therefore be questioned. However, patients seemed to progressively voice greater (if statistically insignificant) satisfaction as more adjunctive modalities were utilized in the informed consent process. Future studies may need to evaluate the balance to be sought between the cost and time demands of such adjuncts against the lack of significant efficacy found in our study. Other adjuncts currently in use include multimedia and patient interactive tools that require similar scrutiny.


All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Gioe had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Johnson, Singh, Gioe.

Acquisition of data. Johnson, Stewart.

Analysis and interpretation of data. Johnson, Singh, Gioe.